consumer psychology of tourism, hospitality and leisure - consumer... · consumer psychology of...

Consumer Psychology of Tourism, Hospitality and Leisure

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page i

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page ii

Consumer Psychology of Tourism,Hospitality and Leisure

Volume 3

Edited by

Geoffrey I. CrouchSchool of Business, Faculty of Law and Management,

La Trobe University, Melbourne, Victoria 3086, Australia

Richard R. PerdueLeeds School of Business, University of Colorado at Boulder,

Boulder, CO 80309-0419, USA

Harry J.P. TimmermansDepartment of Urban Planning, Eindhoven University of Technology,

PO Box 513, 5600 MB, Eindhoven, The Netherlands

Muzaffer UysalDepartment of Hospitality and Tourism Management, Virginia Polytechnic and State

University, 362 Wallace Hall, Blacksburg, VA 24061-0429, USA

CABI Publishing

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page iii

CABI Publishing is a division of CAB International

CABI Publishing CABI PublishingCAB International 875 Massachusetts AvenueWallingford 7th FloorOxfordshire OX10 8DE Cambridge, MA 02139UK USA

Tel: +44 (0)1491 832111 Tel: +1 617 395 4056Fax: +44 (0)1491 833508 Fax: +1 617 354 6875E-mail: [email protected] E-mail: [email protected]: www.cabi-publishing.org

© CAB International 2004. All rights reserved. No part of this publication may be reproduced in any form or by any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners.

A catalogue record for this book is available from the British Library, London, UK.

Library of Congress Cataloging-in-Publication DataConsumer psychology of tourism, hospitality and leisure / edited by

A. G. Woodside … [et al.].p. cm.

Includes bibliographical references.ISBN 0-85199-322-2 (alk. paper)1. Tourism- -Psychological aspects. 2. Travelers - -Psychology.

3. Hospitality industry. 4. Consumer behavior. I. Woodside, ArchG.G155.A1c65 1999338.4�791�0019- -dc21 99-31570

CIP

ISBN 0 85199 749 X

Typeset in 9pt New Baskerville by Columns Design Ltd, ReadingPrinted and bound in the UK by Biddles Ltd, King’s Lynn

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page iv

Contents

Contributors ix

Preface xiii

1 Building Foundations for Understanding the Consumer Psychology of Tourism, 1Hospitality and Leisure

Geoffrey I. Crouch, Richard R. Perdue, Harry J.P. Timmermans and Muzaffer Uysal

PART 1: ATTITUDES, EMOTIONS AND INFORMATION PROCESSING

2 Profiling the One- and Two-star Hotel Guests for Targeted Segmentation Action: a 11Descriptive Investigation of Risk Perceptions, Expectations, Disappointments andInformation Processing Tendencies

Sara Dolnicar

3 The Influence of Consumers’ Emotions on their Service Product Evaluation 21Sandra Gountas and John Y. Gountas

4 Validating a Guttman-type Social Distance Scale for Explaining Residents’ 33Attitudes towards Tourism

Maree Thyne and Andreas H. Zins

PART 2: MOTIVATION AND LEARNING

5 Motivation for Domestic Tourism: a Case Study of the Kingdom of Saudi Arabia 51Naima B. Bogari, Geoff Crowther and Norman Marr

6 Ecotourists’ Environmental Learning Opportunity as a Source of Competitive 65Advantage: Are Ecotourism Operators Missing the Boat with their Advertising?

Garry G. Price

PART 3: CONSUMPTION SYSTEMS

7 Domestic Leisure Traveller Consumption Systems 75Elizabeth Cowley, Ray Spurr, Peter Robins and Arch G. Woodside

8 Tourist Activity Planning in Congested Urban Tourism Environments: Towards 91a Game-theoretic Model and Decision Support System

Qi Han, Benedict G.C. Dellaert, W. Fred van Raaij and Harry J.P. Timmermans

v

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page v

9 Comparing First-time and Repeat Visitors’ Activity Patterns in a Tourism 103Environment

Astrid D.A.M. Kemperman, Chang-Hyeon Joh and Harry J.P. Timmermans

PART 4: DECISION AND CHOICE

10 A Study of Tourist Decision Processes: Algarve, Portugal 121Antónia Correia and Geoffrey I. Crouch

11 The Consumption of Association Convention Sites: Preliminary Results from 135a Study of Site Choice

Geoffrey I. Crouch and Jordan J. Louviere

12 Context and Dynamics of Social Interaction and Information Search in Decision 149Making for Discretionary Travel

Tzung-Cheng Huan and Jay Beaman

13 A Duality in Vacation Decision Making 161Kenneth F. Hyde

14 A Model of Vacation Choice: an Integration of Personality and Vacation Choice 169with Leisure Constraints Theory

Robyn L. McGuiggan

15 Effects of Holiday Packaging on Tourist Decision Making: Some Preliminary 181Results

Walaiporn Rewtrakunphaiboon and Harmen Oppewal

PART 5: EXPERIENCE AND SATISFACTION

16 An Examination of the Antecedents and Consequences of Customer Satisfaction 189Yuksel Ekinci and Ercan Sirakaya

17 First-time and Repeat Visitors to Orlando, Florida: a Comparative Analysis of 203Destination Satisfaction

Paul Fallon and Peter Schofield

18 Aristotelian Ethical Values Within a Tourism/Hospitality Industry Context 215Glenn F. Ross

19 The Role of Expressive and Instrumental Factors in Measuring Visitor 227Satisfaction

Muzaffer Uysal and John Williams

PART 6: MARKET SEGMENTATION

20 Profiling Airline Web Users: a Segmentation Approach 237Joseph S. Chen and Seyou Jang

21 Towards More Thorough Data-driven Segmentation in Tourism: a Tracking 245Framework for Exploring Segment Development

Sara Dolnicar

22 Sustainable Tourism and Stakeholder Groups: a Case Study of Colorado Ski 253Resort Communities

Richard R. Perdue

vi Contents

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page vi

PART 7: ATTRACTION AND LOYALTY

23 Cultural Determinants of Tourist Intention to Return 265Jeffery M. Caneen

24 Towards the Conceptualization of Tourism Destination Loyalty 275Outi Niininen and Michael Riley

PART 8: IMAGE AND INTERPRETATION

25 Measuring Comparative Performance of Vacation Destinations: Using Tourists’ 285Self-reported Judgements as an Alternative Approach

Metin Kozak

26 Cross-cultural Behaviour Research in Tourism: a Case Study on Destination Image 303Metin Kozak, Enrique Bigné, Ana González and Luisa Andreu

27 Journeys of the Imagination? The Cultural Tour Route Revealed 319Tove Oliver

Index 333

Contents vii

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page vii

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page viii

Contributors

Luisa Andreu, Faculty of Business and Economy Studies, Department of Management andMarketing, University of Valencia, Avda. dels Tarongers s/n, 46022 Valencia, Spain. E-mail:[email protected]

Jay Beaman, Auctor Consulting Associates, Ltd, 465 Andra Ct, Cheyenne, WY 82009, USA.E-mail: [email protected]

Enrique Bigné, Faculty of Business and Economy Studies, Department of Management andMarketing, University of Valencia, Avda dels Tarongers s/n, 46022 Valencia, Spain. E-mail:[email protected]

Naima B. Bogari, King Abdul Aziz University, Faculty of Economics and Administration, Women’sCampus, PO Box 42804 Jeddah 21551, Saudi Arabia. E-mail: [email protected]

Jeffery M. Caneen, Brigham Young University, Laie, HI 96762, USA. E-mail:[email protected]

Joseph S. Chen, Department of Hospitality Management, International University of AppliedSciences, Mülheimer Strasse 38, D-53604 Bad Honnef, Germany. E-mail: [email protected]

Antónia Correia, Faculty of Economics, University of Algarve, Campus de Gambelas, 8000-117Faro, Portugal. E-mail: [email protected]

Elizabeth Cowley, School of Marketing, University of New South Wales, Sydney, New SouthWales 2052, Australia. E-mail: [email protected]

Geoffrey I. Crouch, School of Business, Faculty of Law and Management, La Trobe University,Melbourne, Victoria 3086, Australia. E-mail: [email protected]

Geoff Crowther, Department of Marketing, University of Huddersfield, Queensgate,Huddersfield, W. Yorkshire HD1 3DH, UK. E-mail: [email protected]

Benedict G.C. Dellaert, Department of Marketing, Faculty of Economics and BusinessAdministration, Maastricht University, Maastricht, The Netherlands. E-mail:[email protected]

Sara Dolnicar, School of Management, Marketing & Employment Relations, University ofWollongong, Northfields Avenue, Wollongong, New South Wales 2522, Australia. E-mail:[email protected]

Yuksel Ekinci, School of Management, University of Surrey, Guildford, Surrey GU2 7XH, UK.E-mail: [email protected]

Paul Fallon, School of Leisure, Hospitality and Food Management, University of Salford,Frederick Road, Salford M6 6PU, UK. E-mail: [email protected]

Ana González, Faculty of Business and Economy Studies, Department of Management andMarketing, University of León, Campus de Vegazana s/n, 24071 León, Spain. E-mail:[email protected]

ix

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page ix

Sandra Gountas, Department of Marketing, Monash University, Melbourne, Victoria, Australia.E-mail: [email protected]

John Y. Gountas, Bowater School, Deakin University, Geelong, Victoria, Australia. E-mail:[email protected]

Qi Han, Department of Urban Planning, Faculty of Architecture, Building and Planning,Eindhoven University of Technology, PO Box 513, 5600 MB, Eindhoven, TheNetherlands. E-mail: [email protected]

Tzung-Cheng Huan, Graduate Institute of Management, National Chia-yi University, 151 Lin-Sen East Road, Chia-yi, Taiwan, R.O.C. 600. E-mail: [email protected]

Kenneth F. Hyde, Manukau Institute of Technology, Private Bag 94-006, Auckland, NewZealand. E-mail: [email protected]

Seyou Jang, School of Tourism, Sejong University, 98 Gunja-dong, Gwangjin-ku Seoul 143-747,South Korea. E-mail: [email protected]

Chang-Hyeon Joh, Department of Urban Planning, Eindhoven University of Technology, POBox 513, 5600 MB, Eindhoven, The Netherlands.

Astrid D.A.M. Kemperman, Department of Urban Planning, Eindhoven University ofTechnology, PO Box 513, 5600 MB, Eindhoven, The Netherlands. E-mail:[email protected]

Metin Kozak, School of Tourism and Hotel Management, Mugla University, 48000 Mugla,Turkey. E-mail: [email protected]

Jordan J. Louviere, School of Marketing, University of Technology, Sydney, PO Box 123,Broadway, New South Wales 2007, Australia. E-mail: [email protected]

Norman Marr, Department of Marketing, University of Huddersfield, Queensgate,Huddersfield, W. Yorkshire HD1 3DH, UK. E-mail: [email protected]

Robyn L. McGuiggan, Sydney Graduate School of Management, Parramatta, New South Wales2150, Australia. E-mail: [email protected]

Outi Niininen, School of Management, University of Surrey, Guildford, Surrey GU2 7XH, UK.E-mail: [email protected]

Tove Oliver, Institute of Rural Studies, University of Wales, Aberystwyth SY23 3AL, UK. E-mail:[email protected]

Harmen Oppewal, Department of Marketing, Monash University, PO Box 197, Caulfield East,Victoria 3145, Australia. E-mail: [email protected]

Richard R. Perdue, Leeds School of Business, University of Colorado at Boulder, Boulder, CO80309-0419, USA. E-mail: [email protected]

Garry G. Price, School of Tourism and Hospitality, La Trobe University, Bundoora,Melbourne, Victoria 3086, Australia. E-mail: [email protected]

Walaiporn Rewtrakunphaiboon, School of Management, University of Surrey, Guildford,Surrey GU2 7XH, UK. E-mail: [email protected]

Michael Riley, School of Management, University of Surrey, Guildford, Surrey GU2 7XH, UK.E-mail: [email protected]

Peter Robins, Bureau of Tourism Research, GPO Box 1545, Canberra, ACT 2601, Australia.E-mail: [email protected]

Glenn F. Ross, School of Business, James Cook University, Cairns, Queensland 4870, Australia.E-mail: [email protected]

Peter Schofield, School of Leisure, Hospitality and Food Management, University of Salford,Frederick Road, Salford M6 6PU, UK. E-mail: [email protected]

Ercan Sirakaya, Texas A&M University, 256A Francis Hall, 2261 TAMU, College Station, TX77843-2261, USA. E-mail: [email protected]

Ray Spurr, School of Marketing, University of New South Wales, Sydney, New South Wales2052, Australia. E-mail: [email protected]

Maree Thyne, Scottish Centre of Tourism, Aberdeen Business School, The Robert GordonUniversity, Garthdee II, Garthdee Road, Aberdeen AB10 7QG, UK. E-mail: [email protected]

x Contributors

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page x

Harry J.P. Timmermans, Department of Urban Planning, Eindhoven University of Technology,PO Box 513, 5600 MB, Eindhoven, The Netherlands. E-mail: [email protected]

Muzaffer Uysal, Department of Hospitality and Tourism Management, Virginia PolytechnicInstitute and State University, 355 Wallace Hall (0429), Blacksburg, VA 24061-0429, USA.E-mail: [email protected]

W. Fred van Raaij, Department of Economic and Social Psychology, Faculty of Social andBehavioural Sciences, Tilburg University, Tilburg, The Netherlands. E-mail:[email protected]

John Williams, Department of Hotel, Restaurant, Institution Management & Dietetics, KansasState University, 103 Justin Hall, Manhattan, KS 66506-1404, USA. E-mail:[email protected]

Arch G. Woodside, Carroll School of Management, Boston College, 450 Fulton Hall, 140Commonwealth Avenue, Chestnut Hill, MA 02467-3808, USA. E-mail: [email protected]

Andreas H. Zins, Institute for Tourism and Leisure Studies, Vienna University of Economicsand Business Administration, Augasse 2–6, A-1090 Vienna, Austria. E-mail: [email protected]

Contributors xi

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page xi

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page xii

Preface

In Melbourne, Australia, during 5–8 January 2003, a very successful Third Symposium on theConsumer Psychology of Tourism, Hospitality and Leisure (CPTHL) was held and hosted at LaTrobe University. This followed successful symposia held in Hilo, Hawaii (August 1998), andVienna, Austria (July 2000). In Melbourne, 26 papers, selected on the basis of a competitivepaper review process by four reviewers, were presented spanning a broad variety of differentCPTHL topics. Based upon the discussions that took place in Melbourne, authors revised theirpapers before submitting them for review one further time prior to publication in this volume.

The symposium brought together tourism, hospitality and leisure researchers from aroundthe world to report research, share ideas, and advance consumer psychology and consumerbehaviour theory in this important economic sector. In this sector, consumption is character-ized by almost infinite choice, multi-stage and high-involvement decision processes, frag-mented service encounters, multi-dimensional product attributes, globally diverse consumers,and complex social, cultural, ecological and political environments. As such, the tailored devel-opment and application of consumer psychology and consumer behaviour theory to tourism,hospitality and leisure presents interesting challenges and exciting possibilities.

The success of the Third Symposium was due to the efforts of a number of people. First, I amvery grateful to my colleagues and Symposium Co-chairs: Professor Richard Purdue, Professor ofTourism Management at the University of Colorado at Boulder; Professor Harry Timmermans,Professor of Urban Planning at Eindhoven University of Technology in The Netherlands; andProfessor Muzaffer Uysal, Professor of Tourism Management at Virginia Tech. Their input to thesymposium, and their efforts in evaluating papers, chairing sessions and assisting authors with thecompletion of their papers, was critical to the success of the symposium. It was a joy and honourto work with such highly respected international figures in this field. I would also like to acknowl-edge the work of Richard Mitchell, Lisa Damevski and other staff in the Faculty of Law andManagement at La Trobe University who assisted with some aspects of the Symposium and itsorganization. In particular, Megan Morrow provided important support. Rebecca Stubbs at CABIPublishing was most helpful throughout the whole process of publication. All of the authors con-tributed significantly to the success of the Symposium through both their scholarly and socialinput. I would finally like to thank my wife, Linda, and children, Amanda, Vanessa and Scott, fortheir patience and loving support, which has made the task of organizing the Symposium andediting this book that much more satisfying and rewarding.

xiii

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page xiii

Montreal 2005

The Fourth Symposium on the Consumer Psychology of Tourism, Hospitality and Leisure is tobe held in July 2005, in Montreal, Canada, organized by Professor Arch G. Woodside, Professorof Marketing, Carroll School of Management, Boston College and hosted at the École desHEC/HEC School of Management. Further details on the symposium will be available by con-tacting Professor Woodside at [email protected].

Geoffrey I. CrouchMelbourne, March 2003

xiv Preface

Consumer Psych - Chap 00 Prelim 4/12/03 4:20 pm Page xiv

Chapter oneBuilding Foundations for Understanding the ConsumerPsychology of Tourism, Hospitality and Leisure

Geoffrey I. Crouch,1 Richard R. Perdue,2 Harry J.P. Timmermans3 and Muzaffer Uysal41School of Business, La Trobe University, Melbourne, Victoria 3086, Australia; 2Leeds School ofBusiness, University of Colorado at Boulder, Boulder, CO 80309-0419, USA; 3Department ofUrban Planning, Eindhoven University of Technology, PO Box 513, 5600 MB, Eindhoven,The Netherlands; 4Department of Hospitality and Tourism Management, Virginia PolytechnicInstitute and State University, Blacksburg, VA 24061-0429, USA

Introduction

The field of consumer research generally, andconsumer psychology and behaviour morespecifically, has, in this age of consumption,attracted growing interest from marketingscholars and practitioners, psychologists, socialscientists, government policy makers, competi-tion and consumer regulators, consumer advo-cacy groups, and of course consumersthemselves – the general public. A knowledgeof how consumers think, feel, and behave isthe crux of this growing interest since itenables the description, analysis, prediction,and control or influence of consumers, con-sumer systems, and the consumption environ-ment. Both positive and normative aims arebehind the interest in consumers and theresearch that this interest is stimulating. A pos-itive orientation seeks to examine and under-stand what is, whereas a normative approach isaimed at determining what ought to be.

The industries and activities that comprisethe tourism, hospitality and leisure sectors ofthe economy represent one of the largestand fastest growing segments of consumerspending. Yet, despite a developing body ofresearch, our knowledge and understandingof the consumer psychology of tourism, hos-pitality and leisure (CPTHL) is not yet com-mensurate with the economic and socialsignificance of the phenomenon. Indeed,CPTHL research is embryonic and is some-what scattered across many fields of studyand their respective scholarly research jour-nals. However, as noted by Woodside (2000,p. 1), there is growing evidence of the emer-gence of CPTHL research as a separate scien-tific field in terms of academic journals,university degree programmes and scholarlyconferences, including the symposium thathas led to this volume and the earlier publi-cations in this series (Woodside et al., 2000;Mazanec et al., 2001).

© CAB International 2004. Consumer Psychology of Tourism, Hospitality and Leisure, Volume 3 (G.I. Crouch, R.R. Perdue, H.J.P. Timmermans and M. Uysal) 1

Consumer Psych - Chap 01 16/12/03 2:11 pm Page 1

We are therefore now at a stage when thefoundations for developing an understandingof CPTHL are being formed, and the sound-ness of these foundations will shape the pathof future research, and the discoveries andsuccesses it produces.

Dimensions of Psychology

To put CPTHL research into context, we firstconsider the various elements that form thepsychology discipline. The field of psychology,at its most basic, is concerned with understand-ing human behaviour. Since many factors caninfluence, shape or drive human behaviour ina wide variety of contexts, the discipline of psy-chology is quite broad. One way of organizingthis complexity is to sort the field into threedimensions as illustrated in Fig. 1.1.

The first dimension concerns the princi-pal branches of psychology. Examples ofthese overlapping and interconnected fieldsinclude the following:

● Cognitive psychology – a classic definition ofcognitive psychology describes it as refer-ring ‘to all processes by which the sensory

input is transformed, reduced, elaborated,stored, recovered, and used’ (Neiser,1967). Hence cognitive psychology dealswith topics such as perception (attentionand recognition), memory (encoding andstoring, retrieving and forgetting, theorganization of knowledge, and informa-tion processing), language (linguisticknowledge, reading and writing), andthinking, (reasoning, problem-solving andintelligence).

● Environmental psychology – concerns therelationship between human behaviourand the physical environment. Thisbranch of psychology addresses the per-ception and cognition of natural and builtenvironments (Bell et al., 1990). It dealswith issues such as arousal, stimulation,stress, adaptation, approach–avoidancebehaviour, environmental design, way-find-ing, and work versus leisure environments.

● Social psychology – this branch is an amal-gam of sociology and psychology. Itaddresses the ways in which individuals areinfluenced in their behaviour by otherindividuals or groups. Hence, social psy-chology is concerned with social status orclass, peer groups, cultural norms and

2 G.I. Crouch et al.

BRANCH

Cognitive psychology

Social psychology

Environmental psychology

Economic psychology

etc.

CONTEXT

Sport psychology

Child psychology

Forensic psychology

Organizational psychology

Consumer psychology

etc.

LEVEL OFANALYSIS

Sense

Think

Feel

Behave

Cognition

Affect

Conation

}

Fig. 1.1. Dimensions of psychology.


social rules, family and reference groups,etc. It also draws upon the field of culturalanthropology.

● Economic psychology – this interdisciplinarybranch of psychology links cognitive andsocial psychology with economics.Warneryd (1988, in Crotts and van Raaij,1994) defines it as a ‘discipline that stud-ies the social–psychological mechanismsthat underlie the consumption of prod-ucts and services and other economicbehaviour. It deals with consumer prefer-ences, choices, decisions and factors influ-encing these behaviours as well as theconsequences of decisions and choices inthe satisfaction of needs.’

The level of analysis in Fig. 1.1 constitutes thesecond dimension. In psychology, one mightdifferentiate between how individuals sense,think, feel and behave. At the level of thesenses, psychology focuses on biology andphysiology in order to understand the physicalworkings of the body and the brain. In termsof thinking, psychology stresses cognition,learning, involvement, reasoning and intelli-gence. To understand feelings, psychologyaddresses motivations, attitudes, personality,emotions, moods, beliefs and the role ofaffect. Finally, to understand human behav-iour, psychology examines the influence androle of lifestyle, intentions, persuasion, deci-sion making, choice and satisfaction, amongother factors.

A third dimension recognizes differentcontexts in which psychology might be stud-ied. For example, one might study sports psy-chology (how can athletes, and sportsmenand women improve their performancethrough the application of psychologicalprinciples), child psychology (how can thebehaviour of children be understood andmanaged), forensic psychology (how psychol-ogy can assist in solving legal problems), andorganizational psychology (how can an orga-nization such as an enterprise get the bestperformance from its employees). In thisbook we are, of course, interested in anothercontext: that of consumer psychology specifi-cally in tourism, hospitality and leisure (THL)settings. Thus, we are interested in under-standing how the principles and theoriesfrom cognitive, environmental, social and

economic psychology inform an understand-ing of the behaviour of THL consumers. Assuch, knowledge of consumer psychology isfundamental to the successful development,management, and marketing of THL environ-ments and businesses.

The Consumer Psychology of Tourism,Hospitality and Leisure

Mullen and Johnson (1990, p. 1) define con-sumer psychology ‘as the scientific study ofthe behaviour of consumers’. Foxall (1990)and Foxall and Goldsmith (1994) emphasizethe significance of the cognitive consumer as abasis for this scientific understanding of con-sumer behaviour in that ‘consumer choice isportrayed as an ego-involving sequence ofcognitive, affective, and conative changeswhich precede and predetermine the pur-chase/no purchase outcome’ (Foxall 1990,p. 9). Figure 1.2 illustrates a conceptualmodel of this sequence.

A common way of conceptualizing con-sumer psychology is to consider the three pri-mary stages of the consumption or purchaseprocess: pre-purchase, purchase and acquisi-tion, and post-purchase (Arnould et al.,2002). In each of these stages, THL behav-iour is unique. Specifically, when comparedto most retail products, the THL pre-pur-chase stage tends to occur much further inadvance, frequently involves making purchasedecisions from great distances, and involvesmaking decisions between intangible, highlysymbolic alternatives. The choice of destina-tion(s) and the mix of activities to be under-taken during the vacation involves a complexand time-consuming set of decisions tochoose and assemble the desired product priorto departure. Indeed, many tourists derive aconsiderable part of the pleasure of theirexperience from these activities leading up totheir trip.

The purchase and acquisition stageincludes the trip itself which generally repre-sents the core benefit for the consumer.Usually a trip is very high in experiential andhedonic characteristics, and is complex interms of the number of individual, interdepen-dent activities that create the overall vacation

Building Foundations for Understanding 3


experience. Furthermore, the consumer musttravel, typically with a group of other people,to the ‘factory’ (the THL setting) to consumethe product and is heavily involved in a highlycustomized production process. The consump-tion process may or may not involve the use ofguides and other travel assistance.

The post-purchase stage is also of vitalinterest to THL developers and managers, asmany tourists continue to gain a great dealof pleasure from their purchase after theyreturn home in the form of ‘dinner-tablestories’, the giving of souvenirs, the sharingof photographs with family and friends, andlife-long memories. The post-trip stage alsoinfluences their sense of satisfaction, likeli-hood of returning to the destination, andthe word-of-mouth passed on to other poten-tial visitors. For many consumers, the THLplaces they have visited and the activities inwhich they participate have symbolic valuein the presentation of the self to friends andcolleagues.

Furthermore, as the consumption of THLtakes people out of their normal environ-

ment, the principles of environmental psy-chology have particular relevance. Travellersmust cope with unfamiliar physical and socialenvironments. They must learn how to getaround, and how to do simple things, such asbuying a train ticket, which they take forgranted at home but find more problematicwhen travelling. Strange or different physicalenvironments and visually appealing locationsmay, of course, also represent one of the mostcompelling motivations for visiting a destina-tion. Hence, an understanding of environ-mental psychology can be of great assistanceto tourism practitioners.

Travel also often takes people out of theirown culture, so an understanding of cross-cul-tural behaviour in tourism is important, bothfor the tourists as well as employees intourism industries whose job it is to serve andsatisfy the visitor (Reisinger and Turner,2003). Cross-cultural issues can be the causeof problems and misunderstandings. Theycan also produce some of the most reward-ing, memorable and satisfying experiences ofthe journey.


Stimulussituation

Intention Behaviour

CULTURAL CONTEXT

INTERNAL PROCESSES

CONSUMER

Motivation(desire)

Emotion(feeling)

Learning(association)

Perception(awareness)

Cognition and memory(belief and remembering)

SOCIAL CONTEXT

Fig. 1.2. Elements of consumer psychology. Adapted from Mullen and Johnson (1990).


Thus, the context of THL behaviour isunique. This uniqueness provides both a chal-lenge and an opportunity. The challengeinvolves appropriately applying and under-standing the implications of emergent con-sumer psychology concepts and theories tothis unique context. The opportunity is toextend and contribute to the understandingof consumer psychology both in the THLcontext and in general by focusing on thiscontextual uniqueness.

A substantive body of existing CPTHLresearch has evolved over the past 20 years.More than anyone, Mayo and Jarvis (1981),in their seminal work, promoted the study ofthe psychology of leisure travel by examiningcognitive and affective drivers, and socialinfluences. Their book addressed in turn therole of perception, learning, personality,motivation, attitudes and group influences.Other important contributions to the fieldinclude the selection of papers in Johnsonand Thomas (1992) that span issues such astypologies of tourist roles, motivation, sym-bolic consumption, market segmentation,destination image and satisfaction. Theresearch reported in Uysal (1994) variouslyaddresses visitor expectations, the behaviourof package versus non-package tourists, cross-cultural issues, travel motivation, tourists’knowledge and consumer attitudes. Ross(1998) examined both the individual tourist,in terms of motivation, personality, attitudesand environmental influences, as well as soci-etal and organizational contexts, such as destina-tion images, tourism industry employeefactors and social impacts, in deriving anoverarching model of tourist behaviour thatemphasizes a social network perspective.

Tourist satisfaction was the subject of thebook by Ryan (1995). Satisfaction measure-ment has attracted a great deal of interest inthe business world in recent years and therehas been a significant effort among tourismresearchers to apply this thinking to, ordevelop tourism-specific theories for, theanalysis of tourist satisfaction. A further col-lection of research studies published in Pizamand Mansfield (1999) ranged across con-sumer behaviour topics including destinationchoice and decision-making, informationsearch, consumer loyalties, destination image,

expectations, satisfaction, cognitive distance,perceptual mapping, cross-cultural behaviour,family life cycle influences and lifestyles.Linguistic knowledge and influences – a com-ponent of cognitive psychology – received anin-depth analysis in Dann’s (1996) treatisetitled The Language of Tourism: a SociolinguisticPerspective. Dann contends that, ‘so pervasiveand essential is the language of tourism that,without it, tourism itself would surely cease toexist. In the absence of a sociolinguistic basis,the world’s largest industry would simplygrind to a halt…’ (p. 249). Finally,Swarbrooke and Horner (1999) have pro-duced possibly the first textbook focusedspecifically on consumer behaviour in thecontext of tourism, but which provides lim-ited coverage of the psychological basis ofconsumer behaviour.

This growing body of literature is helpingto build the foundations for understandingthe consumer psychology of tourism, hospital-ity and leisure. Through this book and theassociated symposium, it has been our aim tocontribute to progress in this direction.

Structure and Overview of the Book

The following chapters have been groupedaccording to one or more common themes orcharacteristics. The chapters cover a broadrange of consumer psychology topics and avariety of international contexts and back-grounds. Each chapter is the product of schol-arly research by authors who are diverse notonly in geographical origin but also in theiracademic backgrounds and settings. Theauthors share an interest in understandinghow consumers think, feel, behave, and reactto the uniqueness of the THL experience.

Part 1 – Attitudes, Emotions andInformation Processing

Chapter 2: Risk perceptions, expectations,disappointments and information processing

tendencies

This chapter by Sara Dolni�ar examined thechoice behaviour of guests staying in one- and



two-star Austrian hotels. Hotels in this classtypically do little research to understand theircustomer. The chapter investigates customerbehaviour in comparison with tourists stayingin higher-class hotels.

Chapter 3: Consumer emotions and theirevaluation of service encounters

The complex role that consumer emotionsplay in influencing the evaluation of an air-line experience is the subject of this chapterby Sandra and John Gountas. The study con-cludes that leisure airline passengers’ emo-tions prior to and during a flight are relatedto service provision and the passengers’ over-all satisfaction ratings.

Chapter 4: Social distance and residents’attitudes towards tourism

Maree Thyne and Andreas Zins employed theGuttman scaling method to investigate resi-dent attitudes towards tourism in NewZealand and Austria as a function of socialdistance. The scale was found to work wellbut produced more fruitful results in the caseof Austria. The chapter suggests improve-ments so that this scaling method for attitudemeasurement provides enhanced results.

Part 2 – Motivation and Learning

Chapter 5: Motivation for domestic tourism

This chapter investigates the role of ‘push’ and‘pull’ consumer motivations in the domestictourism industry of Saudi Arabia. Suchresearch in developing countries and Islamiccultures has received little research attentionin the past. Bogari, Crowther and Marr findthat religion and culture play a critical role.

Chapter 6: Environmental learning by eco-tourists

Chapter 6 examines the extent to which envi-ronmental education and learning forms a

substantive part of the operation and market-ing of ecotourism operators in Australia.Based on a content analysis of advertisingmaterial, Price finds that environmental edu-cation plays a minor role and that operatorswho capitalize on this shortcoming mayachieve a competitive advantage.

Part 3 – Consumption Systems

Chapter 7: Domestic leisure travel purchaseand consumption systems

Based on an analysis of unit-record data fromthe Australian National Visitor Survey, thischapter by Cowley, Spurr, Robins andWoodside investigates the potential applica-tion of Quick Clustering to construct cogni-tive maps of tourist consumption behaviour.They show that the method has the potentialto reveal important relationships that mightotherwise be missed in other analyticalapproaches.

Chapter 8: Tourist activity planning incongested urban tourism environments

Han, Dellaert, van Raaij and Timmermanscombine an activity-based approach toanalysing tourists’ behaviour with a game-theoretic methodology to study the interactionbetween tourists, and tourism informationoffices. They assess the value of this approachas part of an Internet-based tourist informa-tion system.

Chapter 9: First-time and repeat visitoractivity patterns

In this chapter, first-time and repeat visitoractivity patterns to a theme park are analysed.Kemperman, Joh and Timmermans find thatthese two groups of customers differ particu-larly in terms of the order of activities fol-lowed. Results from this type of analysis havethe potential to enhance the effectivenessand efficiency of theme park operations.



Part 4 – Decision and Choice

Chapter 10: Tourist perceptions andmotivations

Tourist consumer behaviour in the Algarve,Portugal, is examined with respect to the rela-tionship between pre-decision, decision andpost-purchasing consumer behaviour. Throughfactor analysis, Correia and Crouch reveal theprincipal images and motivations that drivetourism to this region.

Chapter 11: Choice and consumption ofassociation convention sites

Crouch and Louviere employ discrete choicemodelling to assess the factors which influ-ence the selection of host destinations forassociation conventions. Through the designof a choice experiment they find strong evi-dence that convention planners in theAustralian domestic conventions industrytake into consideration a number of impor-tant variables or site attributes.

Chapter 12: Social interaction andinformation search in travel decision making

The dynamics of information search andsocial interaction are investigated by Huanand Beaman. They note that this element oftravel decision making has received littleattention to date, but they argue that bettermodels and better data collection and analy-sis will follow from improvements in the wayin which these factors are understood.

Chapter 13: Vacation decision making

A vacation is, in part, planned in advance, butmany travel decisions are made on tour. Hydeinvestigates this duality in trip decision mak-ing and finds that pre-trip planning is moredeliberate and purposeful (cognitive),whereas decision making that occurs on touris more impulsive (hedonic).

Chapter 14: Personality and leisureconstraints in vacation choice

In this chapter, McGuiggan argues that vaca-tion choice models have typically ignored therole of individual personality, and constraintsor barriers to participation in travel. On thisbasis, a conceptual model, including severalresearch propositions, is formulated and putforward for testing, contending that themodel ought to produce stronger relation-ships between personality and tourist choice.

Chapter 15: Effects of holiday packaging onconsideration and choice

Using an experimental approach,Rewtrakunphaiboon and Oppewal study theeffects of holiday packaging on the likelihoodof destination consideration and intention tovisit. They find statistically that packaging canhave a positive impact on choice, and showthat the experimental approach offers impor-tant advantages with respect to the design ofpackage attributes.

Part 5 – Experience and Satisfaction

Chapter 16: Antecedents and consequencesof customer satisfaction

Customer satisfaction has become a majorresearch theme in marketing generally, andin this chapter Ekinci and Sirakaya investigatecustomer satisfaction in the context of arestaurant setting. One of the main debatesin this area of research concerns the order ofthe causal chain linkages connecting satisfac-tion, quality, attitude and behaviour. Ekinciand Sirakaya contribute to this debate.

Chapter 17: Destination satisfaction

The determinants of first-time and repeat-visi-tor satisfaction of visitors to Orlando, Florida,are investigated in this study by Fallon andSchofield. Using principal component analy-sis, they find that first-time visitor satisfactionwas influenced primarily by accommodation



and hospitality factors, whereas repeat-visitorsatisfaction was governed more by secondaryattractions such as shopping and restaurants.

Chapter 18: Visitor, staff and managementethical beliefs and expectations

Ross investigates, in this chapter, the role ofethical beliefs and values on staff–visitor inter-actions and the overall level of success of theorganization. He looks at the ethical values oftourism industry employees and how thesecompare to the ethical expectations of man-agement and of visitors.

Chapter 19: Expressive and instrumentalfactors in measuring visitor satisfaction

The objective of the research reported in thischapter by Uysal and Williams was to investi-gate whether visitor satisfaction could be pre-dicted on the basis of both instrumentalfactors (i.e. functional factors such as conve-nience, accessibility and quality) and expres-sive factors (i.e. affective constructs such asbeauty, scenic wonders, quaintness, etc.).Their study found that motivation moderatesthe importance of these two groups of factors.

Part 6 – Market Segmentation

Chapter 20: Profiling airline web users

There is today a great deal of interest in the useof customer websites in marketing. Chen andJang contribute to this field in a study of airlineweb users that examines preferred websiteattributes as determinants of market segmenta-tion. They find two major market segments,which they label Bargain Seekers and Utilitarians.

Chapter 21: Data-driven market segmenttrends

This study by Dolni�ar considers some of theadvantages of data-driven, or a posteriori,approaches to market segmentation. Becausedata-driven market segmentation derives mar-

ket segments using criteria that cannot be pre-determined ahead of the trip, the tracking ofsuch segments is problematic. This chaptersuggest how this problem might be addressed.

Chapter 22: Sustainable tourism andstakeholder groups

A detailed case study of Colorado ski communi-ties is presented by Perdue in Chapter 22.Three resort community populations, guests,resort employees and residents, are studied toreveal eight stakeholder groups. The behaviourand attitude of these groups is analysed usingquantitative and qualitative methodologies.

Part 7 – Attraction and Loyalty

Chapter 23: Cultural determinants of touristdestination loyalty

By comparing US, Japanese and Chinesetourists in Hawaii, Caneen investigates, in thischapter, the effect of culture and nationalityon intention-to-return. He finds distinct dif-ferences and suggests a number of implica-tions for destination marketing.

Chapter 24: Conceptualization of tourismdestination loyalty

Niininen and Riley, in this chapter, seek tobuild a stronger foundation for understand-ing and analysing the concept of destinationloyalty. To this end, they develop a conceptualmodel of destination loyalty, which recognizesthe role of personality, optimum stimulationlevel, attitudes toward repetitive holidays,vacation barriers and repeat behaviour.

Part 8 – Image and Interpretation

Chapter 25: Tourists’ judgements inmeasuring comparative destination

performance

In a study of British tourists visiting Turkeyand Mallorca, Kozak tests an approach



designed to measure the performance ofinternational tourist destinations in terms oftheir perceived competitive position. He con-cludes that self-reported judgements can helpto diagnose competitive positions with impli-cations for destination management and mar-keting.

Chapter 26: Cross-cultural behaviourresearch on destination image

Kozak used analysis of variance and factorcorrespondence analysis to analyse differ-ences in destination image attributes ofValencia, Spain, in a survey of visitors fromBelgium, Britain, France, Germany and TheNetherlands. Kozak et al. show that imagesvary significantly and that important insightsinto the motivations for travel by differentcultures can assist in the development of des-tination marketing strategies.

Chapter 27: Journeys of the imagination: acultural tour route

Using an environmental psychology perspec-tive, Oliver employed a route-mapping tech-nique to elicit information about tourists’knowledge of an organized group tour routebefore and after the trip. Her qualitative andquantitative approach demonstrates howchanges occur in this knowledge, particularlyin its magnitude. The chapter emphasizes thevalue of research focuses on destination-

image formation during the critical pre-pur-chase period.

Conclusions

CPTHL is a fertile field for research. As theglobal tourism industry has grown, interest inunderstanding the tourism consumer hasbecome paramount to the success of tourismenterprises, travel destinations and host com-munities. It is also a vital input to tourism pol-icy, planning for sustainable tourismdevelopment and government programmesfor economic development and destinationmarketing. Yet the bulk of tourism consumerstudies focus on tourists en masse. They tendto emphasize aggregate tourism statistics andthe determinants of demand, with particularfocus on demographic, macro-economic andsocio-cultural trends. A minority of this con-sumer research attempts to understand thepsychology of the tourist or of the host com-munity residents.

Mayo and Jarvis (1981), who pioneeredthe study of CPTHL, argued that ‘by focusingon the individual rather than the “average” –and by understanding this traveller in psy-chological instead of merely demographicterms – new insights into travel behaviourwill be possible’ (p. xiii). Furthermore,because of its uniqueness, insights into travelbehaviour will also contribute to the moregeneral understanding of consumer behav-iour. The following chapters in this bookcontinue this quest.


References

Arnould, E., Price, L. and Zinkhan, G. (2002) Consumers, 1st edn. McGraw Hill, New York.Bell, P.A., Fisher, J.D., Baum, A. and Greene, T.C. (1990) Environmental Psychology. Harcourt Brace

Jovanovich, Orlando, Florida.Crotts, J.C. and van Raaij, W.F. (eds) (1994) Economic Psychology of Travel and Tourism. The Haworth Press,

Binghamton, New York.Dann, G.M.S. (1996) The Language of Tourism: a Sociolinguistic Perspective. CAB International, Wallingford,

UK.Foxall, G.R. (1990) Consumer Psychology in Behavioural Perspective. Routledge, London.Foxall, G.R. and Goldsmith, R.E. (1994) Consumer Psychology for Marketing. Routledge, London.Johnson, P. and Thomas, B. (eds) (1992) Choice and Demand in Tourism. Mansell Publishing, London.Mayo, E.J., Jr and Jarvis, L.P. (1981) The Psychology of Leisure Travel. CBI Publishing, Boston, Massachusetts.Mazanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) (2001) Consumer Psychology of Tourism,

Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK.


Mullen, B. and Johnson, C. (1990) The Psychology of Consumer Behavior. Lawrence Erlbaum Associates,Hillsdale, New Jersey.

Neiser, U. (1967) Cognitive Psychology. Appleton-Century-Crofts, New York.Pizam, A. and Mansfeld, Y. (eds) (1999) Consumer Behavior in Travel and Tourism. The Haworth Hospitality

Press, Binghamton, New York.Reisinger, Y. and Turner, L.W. (2003) Cross-cultural Behaviour in Tourism: Concepts and Analysis.

Butterworth–Heinemann, Oxford.Ross, G.F. (1998) The Psychology of Tourism, 2nd edn. Hospitality Press, Melbourne.Ryan, C. (1995) Researching Tourist Satisfaction: Issues, Concepts, Problems. Routledge, London.Swarbrooke, J. and Horner, S. (1999) Consumer Behaviour in Tourism. Butterworth–Heinemann, Oxford.Uysal, M. (ed.) (1994) Global Tourist Behavior. International Business Press, Binghamton, New York.Warneryd, K.E. (1988) The psychology of innovative entrepreneurship. In: van Raaij, W.F., van Veldhoven,

G.M. and Warneryd, K.E. (eds) Handbook of Economic Psychology. Kluwer Academic, Dordrecht.Woodside, A.G. (2000) Introduction: theory and research on the consumer psychology of tourism, hospi-

tality and leisure. In: Woodside, A.G., Crouch, G.I., Mazanec, J.A., Oppermann, M. and Sakai, M.Y.(eds) Consumer Psychology of Tourism, Hospitality and Leisure. CAB International, Wallingford, UK,pp. 1–17.

Woodside, A.G., Crouch, G.I., Mazanec, J.A., Oppermann, M. and Sakai, M.Y. (eds) (2000) ConsumerPsychology of Tourism, Hospitality and Leisure. CAB International, Wallingford, UK.



Chapter twoProfiling the One- and Two-star Hotel Guests for TargetedSegmentation Action: a Descriptive Investigation of RiskPerceptions, Expectations, Disappointments and InformationProcessing Tendencies

Sara Dolnic�arSchool of Management, Marketing and Employment Relations,University of Wollongong,Wollongong, NSW 2522, Australia

Abstract

Identifying the target segment is the basis of developing efficient market segmentation strategies, and effi-cient market segmentation is vital in an industry that is becoming increasingly competitive, as in the caseof international tourism. In Austria, hotels in higher-star grading categories have addressed this needthrough systematic market research designed to identify the needs of their consumers. Not so the hotels inthe one- and two-star category: these typically do not segment the market and tend to assume that increas-ing their star grading will lead to increased market demand instead of investigating the specific needs oftourists who very consciously choose low-star graded hotels. This chapter aims to examine this a priori seg-ment with regard to issues that are known to influence choice behaviour, namely expectations, disappoint-ments with past experiences and perceived risk, while taking into account information need andprocessing habits. The ultimate purpose of the study is to stimulate development of a segment-orientedmarketing strategy for one- and two-star hotels should this guest segment differ significantly from thatcomprising tourists staying in higher-graded hotels.

Introduction and Foundations

Within the tourism industry, a priori marketsegmentation (Mazanec, 2000) has becomethe most widespread strategic marketing con-cept practically applied. Various tourist char-acteristics can be used for this purpose. Fromthe perspective of the accommodation sector,one of the most interesting criteria for group-

ing tourists is the category of hotel chosen.The assumption that guests who frequent thesame hotel category are more homogeneouswith respect to their judgements towardsaccommodation than guests choosing otherhotel categories is the reason that this group-ing criteria is of relevance.

Being aware of the judgements of the seg-ment that is served by the specific business



empowers a hotel to make adaptations (inproduct and service offered, advertising strat-egy, choice of distribution channels, pricingpolicy, etc.) customized to attract and satisfythe market segment targeted. Such matchingof expectations of the target market segmentfavours inclusion into a consumer’s consider-ation set (Woodside and Lysonski, 1989;Moutinho, 2000a) as well as offering thepotential to increase guest satisfactionthrough low deviation from expectations andexperiences (Moutinho, 2000b) and thusgenerating loyalty and repeat business(Decrop, 2001). Such strategies also decreaseconsumer dissatisfaction resulting from theattribution of negative feelings to externalfactors, and leading to attitudinal and behav-ioural change unfavourable to the particularcompany or destination (Decrop, 2001).

Consequently, maximum understandingof the targeted market segment increases theprobability of market success probability,making any attempt to gain insight into the‘mind of the market segment’ more thanworthwhile. The components of the ‘tourists’minds’ that are of particular interest in thisstudy include the kind of information soughtwhen preparing to choose a hotel, the infor-mation processed in the course of makingsuch a decision, the risks perceived whenchoosing an unknown hotel, the expecta-tions with regard to the one- and two-star cat-egory and finally prior disappointmentsexperienced. These factors were chosenbecause they help the tourism industry todetermine the central concerns of the mar-ket segment under consideration as well asdeduce managerial consequences fromthem, as it has been widely shown that thesefactors influence travel and/or travel deci-sion behaviour (the influence of past experi-ences has been shown by Mazursky, 1998;compare references Sönmez and Graefe,1998; King and Woodside, 2001; of perceivedrisk by Goodrich, 1991; Um and Crompton,1990; Roehl and Fesenmaier, 1992; Sönmezand Graefe, 1998; of dis/satisfaction byDecrop, 2001). The information needed forand processed during the actual hotel deci-sion making process is relevant for the com-munication strategy of the hotel. Thus,knowing which information is needed and

processed during the hotel decision makingprocess helps the one- and two-star hotels toinclude relevant information in theirbrochures or on their homepages (communi-cation match with the target segment cho-sen). Risks perceived are extremely useful forhotels in determining communication strat-egy and product design, and expectationsand disappointments allow hotel manage-ment to minimize the expectation–experi-ence gap that is crucial for either asatisfactory experience potentially leading torepeat visits and loyalty, or perceived dissatis-faction leading to negative attitudes andbehavioural change toward the hotel andmaybe even demotion to one- and two-starhotel status.

The reasons for investigating the a priorisegment of guests staying in one- and two-starhotels are twofold: (i) all the issues men-tioned so far have widely been studied bothin consumer behaviour literature as well as inthe field of tourism research. However, thesetopics and the consequences for the tourismindustry resulting from these findings have –to the author’s knowledge – never focused onthe segment of hotel guests staying in low-starcategory hotels. This fact can be well illus-trated in exploring studies that aim to iden-tify the most important hotel attributes fromthe guest perspective. Among 20 studies(Lewis, 1984a,b; Cadotte and Turgeon, 1988;Wind et al., 1989; Saleh and Ryan, 1991,1992; Anath et al., 1992; Barsky and Labagh,1992; McCleary et al., 1993; Pannell KerrForster Associates, 1993; Weaver and Oh,1993; Clow et al., 1994; Schaefer et al., 1995;Tsaur and Tzeng, 1995; Griffin et al., 1996;Gundersen et al., 1996; Hartline and Jones,1996; Bowen and Shoemaker, 1998; Dubeand Renaghan, 1999a,b, 2000a,b) that inessence pursue this goal,1 40% do not studyany particular group of travellers, 25% focuson business travellers, 10% explore bothfour-star hotel guests and intermediaries andone study investigates hotel attribute impor-tance in the three-star, the luxury and themature traveller segments. (ii) In theAustrian lodging industry the finding thatsegmentation is a profitable concept is notgenerally accepted. Systematically, four- andfive-star hotels are well aware of this fact and

12 S. Dolnic�ar


act accordingly, whereas one- and two-starhotels in particular do not lay any impor-tance on segment understanding. This is dueto two main factors: first, hotels in the one-and two-star category do not have organiza-tional structures that allow for strategic man-agement. This can be described as a‘structural problem’ of the lodging industry.Second, there is a tendency for every hotel toaspire towards gaining an additional star.The one- and two-star businesses thus feelinferior within the lodging industry insteadof taking advantage of their market opportu-nities (‘star greediness problem’).

The aim of this chapter is to examine themake-up of visitors to Austria who choose tostay in one- or two-star graded accommoda-tion to improve: (i) understanding of thisgroup and (ii) evaluation of the distinctnessof this group as a target segment. If distinctprofiles emerge, the one- and two-star hotelsuse such information to build up a strongbrand, which consequently would weaken the‘star greediness problem’. The ‘structuralproblem’ could be solved in a second step byproviding central market research and strate-gic marketing support for member hotels.

Exploring the One- and Two-star HotelGuests in Austria

Six hundred and fourteen interviews wereconducted on the basis of hypothesis-ori-ented quota sampling within the followingcategories: season (248 respondents werequestioned during the winter season of 2001,366 in the summer months, the sample sizeof the one- and two-star guest amounting to147), country of origin, city or non-city desti-nation, business or vacation travel purposeand star grading categories. The interviewtook about 15 min and was conducted in thehotels in which the guests stayed with permis-sion of the owners. Each respondent wasinstructed to answer the question only withregard to his or her present travel purposeand for the star category of the hotel atwhich they were staying at that time. Theexact questions and results are provided inthe following subsections.

Pre-choice information search andinformation processing during decision

making

Two questions were posed in order to investi-gate what kind of information is soughtbefore the hotel selection process, and whatkind is processed during decision making.The first question requires the respondents toprovide an answer without support of theinterviewer and is formulated in the follow-ing manner: ‘Think of a hotel decision for avacation/business trip. You have to chooseone of two hotels. Both hotels are unknownto you at this point. Which information doyou collect in order to take this decision?’

The answers were recorded according tothe rank of the item as assigned by therespondent. If a respondent answered thequestion by indicating ‘Price, location andcleanliness’, price would be coded ‘1’, loca-tion ‘2’ and cleanliness ‘3’. Results are pro-vided in Table 2.1. The first column gives thepercentage of respondents choosing eachparticular factor, the median rank is com-puted only on the basis of the respondentsstating the issue. The last two columnsinclude minimum and maximum rank.

The results indicate that price is the factornamed most often (83% give price as rele-vant information and, for most of theserespondents, price is top of the list). Slightlymore than two-thirds name the locationand/or the surroundings of the hotel, almosttwo-thirds the room, the latter two aspectstypically ranked second in the lists of therespondents. The remaining issues arenamed by less than one-third of the respon-dents. An interesting observation is that thestar category is named by only 14% of therespondents, but among these guests, thestar category seems to play an importantrole, as the median rank is 2 and even themaximum rank of 4 is lower than is the casefor the leading three factors.

The second question approaches theissue from a more applied perspective.Respondents are given copies of pages froma hotel guide for two hotels. They are askedto look at the descriptions carefully andmake a decision regarding which one ofthese two hotels they would book. After

Profiling Guests for Targeted Segmentation Action 13


making their decision, the page from thehotel guide is removed and respondents areasked to tell the interviewer which pieces ofinformation they remember, both for the‘booked’ and the ‘rejected’ hotel. The rank-ing of hotel information resulting from thisquestion is provided in Fig. 2.1.

The picture and the star category are fore-most, with 63% and 54% of the respondentsremembering this information for the hotel

they decided to choose in the interview set-ting and 47% each for the hotel rejected inthis process (the chi-squared test does notsupport the hypothesis that the differencesbetween the chosen and the rejected hotelare significant, rendering a P-value of 0.324,although the sum of all pieces of informationdiffers strongly). The name is remembered byhalf of the respondents that booked thehotel, the price by 41%.

14 S. Dolnic�ar

Table 2.1. Unaided statements of pieces of information needed for decision making (rank coded).

% Stated Median rank Minimum rank Maximum rank

Price 83% 1 1 5Location/surroundings 65% 2 1 5Room 56% 2 1 5Food and drink 29% 4 1 9Reachability 27% 3 1 7Staff (friendliness…) 24% 4 1 9Hotel equipment 24% 3 1 8Star category 14% 2 1 4Technical equipment in the room 10% 3 2 5Pool 9% 5 2 8Garages/parking possibility 7% 5 3 8Picture/looks 7% 6 1 8That everything works 7% 5 1 10Sauna 4% 5 3 8Workout room 3% 4 2 6

63%

54%

50%

41%

12%

10%

10%

47%

47%

26%

29%

7%

3%

4%

0% 10% 20% 30% 40% 50% 60% 70%

Picture

Star category

Name

Price

Garage

Restaurant

TV‘booked’ hotel‘rejected’ hotel

Fig. 2.1. Information remembered after the hotel choice process.


Factors of perceived risk

Tackling the issue of relevant hotel factorsfrom a different perspective includes askingguests which aspects they consider to be mostrisky when booking a hotel in a specific starcategory. The question was formulated as fol-lows: ‘There is not a single room available inyour favourite hotel any more. You are there-fore in the situation of having to book aroom in a hotel that you have no prior expe-rience with. What factors are you mostuneasy about?’

One hundred and twenty-three out of the147 respondents in the one- and two-starsample (84%) answered this open-format ques-tion. Among those, 86% listed two or three fac-tors of perceived risk. The frequency table forall statements given by at least 5% of the totalsample (147) is provided in Table 2.2.

Location, price and cleanliness representthe top three risks perceived. Among thoserespondents who stated risks when ques-tioned (123), more than half named ‘loca-tion’, and more than one-third named either

‘price’ and ‘cleanliness’. The concerns of thisgroup of visitors thus strongly centre aroundfundamental product components or ‘hardfacts’, as compared to ‘soft facts’ such as ser-vice quality and friendliness.

One concern dominates the list of per-ceived risks among the one- and two-starguests: location. As ‘location’ implies a widevariety of aspect, respondents were addition-ally asked to answer the following question:

During this particular stay, is it very important,important, not so important or absolutely notimportant …

… that the hotel is easy to reach from theairport or train station?… that the hotel is easy to reach by car?… that the hotel is located near sites impor-tant to you (ski lift, conference centre, etc.)?… that the hotel is close to nature?… that the hotel is located in the centre ofthe city?… that the hotel is located in a quiet region?’

The distribution of answers is provided inTable 2.3. The numbers indicate the percent-age of respondents indicating each particular


Table 2.2. Risks perceived by one- and two-star hotel guests.

Worried concerning … Frequency Percentage of total sample

Location 64 44Price 46 31Cleanliness 41 28Food 21 14Staff/friendliness 20 14Room 14 10Bathroom 12 8Furniture/equipment 12 8Service 11 7Bed 7 5Quietness 7 5

Table 2.3. Importance of location factors.

LocatedEasy to reach Easy to near Located in Located infrom airport/ reach by important Close to centre of a quiettrain station car sites nature the city region

Absolutely not important 14% 23% 7% 12% 16% 3%Not so important 25% 14% 20% 33% 18% 18%Important 20% 30% 44% 31% 37% 43%Very important 40% 32% 29% 24% 30% 35%


agreement level with regard to the locationfactors investigated. It becomes apparentfrom these results that the segment of one-and two-star hotel guests is interested in loca-tions that are easy to reach from the airportor train station. The second most importantcomponent of the location is quietness;approximately one-third of these tourists con-sider it to be very important that the hotelcan comfortably be reached by car and thatthey are close to the centre of the city.

Expectations

The investigation on the expectations of one-and two-star hotel guests was undertakenfrom two perspectives. On the one hand,there was a direct question, aimed at reveal-ing what the visitors expect from ‘their’ usualhotel star category. (‘So you have a lot ofexperience with hotels within the … star cate-gory. What do you expect from …-starhotels?’). On the other hand, respondentswere asked to state in open-question formattheir reasons for dissatisfaction with this par-ticular star category. (‘What were – for youpersonally – the main reasons for dissatisfac-tion in …-star hotels?’).

The answers to the first question (percent-ages provided in Table 2.4) are dominated byone factor: 40% of the respondents expressexpectations in the area of cleanliness orhygiene (both in the hotel and in the room).

The remaining statements are almost negligi-ble in comparison: 7% expect a bathroom, ashower, the location and the service to begood; 6% expect good food, and the priceissue is named by 8% of the respondents, withhalf of them verbalizing the issue as ‘cheap’,half of them emphasizing the ‘value formoney’ perspective.

Disappointments

Past disappointments with hotel experienceswithin these star grading categories very wellmirror the expectations discussed above. Againthe issue of hygiene and cleanliness is namedmost frequently. Contrarily, the price seems tobe more dominant in the minds of tourists stay-ing in one- and two-star hotels before the stayin the form of expectations: only 3% of therespondents state that they have been disap-pointed by the fact that the price was too high.The detailed list of all disappointments (cate-gorized) is provided in Table 2.5.

Contrasting ‘low-’ and ‘high-star-category’guests

A number of significant contrasts comparedto ‘higher’ hotel categories are revealed: withregard to the unaided statement of pieces ofinformation needed for the hotel decision,the star category is stated significantly more

16 S. Dolnic�ar

Table 2.4. Most frequently stated expectations of one- and two-starhotel guests.

Percentage of all Expectation Frequency respondents

Cleanliness/hygiene 54 40Bathroom 11 7Shower 11 7Good location 10 7Good service 10 7Good food 9 6TV 8 5Cheap 6 4Friendly 6 4Comfortable 6 4Good value for money 6 4


often among guests of higher-star categoriesthan among guests staying at one- and two-star places (Pearson chi-squared P-value =0.000 with 6% of the one-star, 16% of the two-star, 28% of the three-star, 34% of the four-star and 38% of the five-star guests activelysearching for this piece of information). Thecontrary is true for the price information(chi-squared P-value of 0.000, about 80% ofall guests staying in the one- and two-star cate-gory ask for the price information, only 67%in the three-star category, followed by 55% inthe four- and 27% in the five-star hotels).Other significant differences include theinquiry whether there is a sauna and a gym inthe hotel (independent questions, P = 0.000for both chi-squared tests) with 19% of thefive-star hotel guests asking for both pieces ofinformation and only less than 10% in allother hotel categories.

The comparison of information remem-bered after the simulated hotel choice duringthe interview reveals that guests staying infive-star hotels pay significantly more atten-tion to hotel features like sauna (P = 0.000)and gym (P = 0.010), whereas far more one-(52%) and two-star guests (37%) rememberthe price (P = 0.030).

Differences with regard to perceived risksare detected in a purely descriptive manner,because the expected cell frequency does notallow for statistical testing: the quality of thebed, the food quality, hygiene, cleanlinessand price seem to be perceived as risky moreoften in the one- and two-star hotel category,whereas quality in general, security and the

star category are the major concerns for guestchoosing high-star categories.

Differences with respect to expectationsmirror perceived risks expressed by respon-dents (again, cell frequency is too low due tothe large number of expectations stated):cleanliness is mentioned most often amongguests choosing the one-star category (17%state this particular issue). Furthermore,guests staying in low categories more oftenname food, the furnishings of the room, loca-tion, TV, minibar, air conditioning and escala-tor, attached bathroom, comfort and goodvalue for money. Guests choosing accommo-dation in high-star graded hotels expect theirstay to be quiet, they count on high quality ingeneral, good food, sauna, gym, perfect ser-vice, good ambience, parking facilities andbusiness infrastructure more often. Theexpectations verbalized by guests from low-starcategories in general are less abstract than inthe case of guests in higher-star categories.This is supported by the differences detectedin the field of prior disappointments: cleanli-ness and food quality have often led to dis-satisfaction among guests in low-star hotelcategories, whereas disappointments in high-star categories – if articulated at all – typicallyconcern intangible human components of theproduct, especially the human factor.

This contrasting perspective indicates thatthe one- and two-star hotel guests do repre-sent a profiled and distinct market segmentthat can very well be targeted by an appropri-ate product and that could be marketed usingthe star grading as a brand system.


Table 2.5. Reasons for disappointments.

Percentage of all Reason for disappointment Frequency respondents

Hygiene/cleanliness 36 24Food 20 14Room 19 13Staff 17 12Bed 10 7Noise 9 6Bathroom 8 5Service 6 4Too expensive 5 3Technical equipment 3 2


Conclusions

The main aim of the study presented was todetermine whether a distinct market segmentexists that could be targeted by the one- andtwo-star category hotels. A data set including614 interviews (147 of which were conductedin one- and two-star hotels) formed theempirical basis for the investigation. Themain limitation of the study is the small sam-ple size within the one- and two-star hotel cat-egory (that resulted from major field workdifficulties due to the small size of the hotelsand the fact that these accommodations arenot open all year). Descriptive analysis of theone- and two-star guest segment renders thefollowing results:

● Price is the one factor stated most often tobe relevant information for choosing ahotel, followed by location and the roomcondition. Among those respondents thatstated the star category, this particularpiece of information is mostly placed inthe second position, thus indicating thatthere is potential for a one- and two-starbranding endeavour.

● The picture (photo printed in the hotelguide) and the star category are remem-bered most often after making the actualhotel choice.

● The top three risks perceived are location,price and cleanliness, where locationmostly indicates reachability and quietness.

● The top expectations include cleanlinessand hygiene, the bathroom, the locationand service.

● Disappointments have mostly beenencountered with regard to hygiene,cleanliness, food and the room in the one-and two-star hotel grading category.

Significant differences between the guestsstaying in one- and two-star hotels and guestsstaying in higher categories could be detectedwith regard to all criteria studied: pieces ofinformation needed for the hotel decision,information remembered after the simulatedhotel choice, perceived risks, expectationsand disappointments, thus supporting theassumption that a distinct market segmentexists that could be targeted by hotels of thisgrading category, providing a better matchbetween demand and supply than can beachieved by trying to upgrade the hotel tohigher-star categories.

Acknowledgement

This study was partially conducted duringSara Dolnicar’s appointment as assistant pro-fessor at the Institute for Tourism and LeisureStudies at the Vienna University ofEconomics and Business Administration andsupported by the FWF grant 010 (AdaptiveInformation Systems and Modeling inEconomics and Management Science). Thedata underlying this study was collected aspart of an empirical research project con-ducted in cooperation with Dr Thomas Otter(Ohio State University, USA) for the AustrianBusiness Chamber and the Austrian FederalMinistry of Economics and Labour.

Note

1Although various perspectives are taken, such asimportance for the choice of a hotel, influence onthe intention to revisit and loyalty generation, cus-tomer satisfaction etc.

18 S. Dolnic�ar

References

Ananth, M., DeMicco, F.J., Moreo, P.J. and Howey, R.M. (1992) Marketplace lodging needs of mature trav-ellers. The Cornell Hotel and Restaurant Quarterly August, 12–24.

Barsky, J.D. and Labagh, R. (1992) A strategy for customer satisfaction. The Cornell Hotel and RestaurantAdministration Quarterly October, 32–40.

Bowen, J.T. and Shoemaker, S. (1998) Loyalty: a strategic commitment. The Cornell Hotel and RestaurantAdministration Quarterly February, 12–25.

Cadotte, E.R. and Turgeon, N. (1988) Key factors in guest satisfaction. The Cornell Hotel and RestaurantAdministration Quarterly February, 45–51.


Clow, K.E., Garretson, J.A. and Kurtz, D.L. (1994) An exploratory study into the purchase decision processused by leisure travellers in hotel selection. Journal of Hospitality and Leisure Marketing 4, 53–71.

Decrop, A. (2001) The antecedents and consequences of vacationers’ dis/satisfaction: tales from the field.In: Mazanec, J.A., Crouch, G.I., Ritchie, J.R. and Woodside, A.G. (eds) Consumer Psychology of Tourism,Vol. 2. CAB International, Wallingford, UK, pp. 333–348.

Dube, L. and Renaghan, L.M. (1999a) Building customer loyalty. The Cornell Hotel and RestaurantAdministration Quarterly October, 78–88.

Dube, L. and Renaghan, L.M. (1999b) How hotel attributes deliver the promised benefits. The Cornell Hoteland Restaurant Administration Quarterly October, 89–95.

Dube, L. and Renaghan, L.M. (2000a) Creating visible customer value – how customers view best-practicechampions. The Cornell Hotel and Restaurant Administration Quarterly February, 62–72.

Dube, L. and Renaghan, L.M. (2000b) Marketing your hotel to and through intermediaries – an over-looked best practice. The Cornell Hotel and Restaurant Administration Quarterly February, 73–83.

Goodrich, J.N. (1991) An American study of tourism marketing: impact of the Persian gulf war. Journal ofTravel Research 30, 37–41.

Griffin, R.K., Shea, L. and Weaver, P. (1996) How business travellers discriminate between mid-priced andluxury hotels: an analysis using a longitudinal sample. Journal of Hospitality and Leisure Marketing 4,63–75.

Gundersen, M.G., Morten, H. and Olsson, U.H. (1996) Hotel guest satisfaction among business travellers.The Cornell Hotel and Restaurant Administration Quarterly April, 72–81.

Hartline, M.D. and Jones, K.C. (1996) Employee performance cues in a hotel service environment: influ-ence on perceived service quality, value, and word-of-mouth intentions. Journal of Business Research 35,207–215.

King, R.L. and Woodside, A.G. (2001) Qualitative comparative analysis of travel and tourism purchase-con-sumption systems. In: Mazanec, J.A., Crouch, G.I., Ritchie, J.R. and Woodside, A.G. (eds) ConsumerPsychology of Tourism, Vol. 2. CAB International, Wallingford, UK, pp. 87–106.

Lewis, R.C. (1984a) Isolating differences in hotel attributes. The Cornell Hotel and Restaurant AdministrationQuarterly November, 64–77.

Lewis, R.C. (1984b) The basis of hotel selection. The Cornell Hotel and Restaurant Administration QuarterlyMay, 54–69.

Mazanec, J. (2000) Market segmentation. In: Jafari, J. (ed.) Encyclopedia of Tourism. Routledge, London, pp.374–375.

Mazursky, D. (1980) Part experience and future tourism decisions. Annals of Tourism Research 16, 333–344.McCleary, K.W., Weaver, P.A. and Hutchinson, J.C. (1993) Hotel selection factors as they relate to business

travel situations. Journal of Travel Research 32, 42–48.Moutinho, L. (2000a) Consumer behaviour. In: Mountinho, L. (2000) Strategic Management in Tourism. CAB

International, Wallingford, UK, pp. 41–78.Moutinho, L. (2000b) Segmentation, targeting, positioning and strategic marketing. In: Mountinho, L.

(2000) Strategic Management in Tourism. CAB International, Wallingford, pp. 41–78.Pannell Kerr Forster Associates (1993) Factors influencing the design of hotels. International Journal of

Contemporary Hospitality Management 5, 17–19.Roehl, W.S. and Fesenmaier, D.R. (1992) Risk perception and pleasure travel: an exploratory analysis.

Journal of Travel Research 30, 17–26.Saleh, F. and Ryan, C. (1991) Analysing service quality in the hospitality industry using the SERVQUAL

model. The Service Industries Journal 11, 324–343.Schaefer, A., Illum, S. and Margavio, T. (1995) The relative importance of hotel attributes to motorcoach

tour operators. Journal of Hospitality and Leisure Marketing 3, 65–80.Sönmez, S.F. and Graefe, A.R. (1998) Determining future travel behaviour from past travel experience and

perceptions of risk and safety. Journal of Travel Research 37, 171–177. Tsaur, S.-H. and Tzeng, G.-H. (1995) Multiattribute decision making analysis for customer preference of

tourist hotels. Journal of Travel and Tourism Marketing 4, 55–69.Um, S. and Crompton, J.L. (1990) Attitude determinants in tourism destination choice. Annals of Tourism

Research 17, 432–448.Weaver, P.A. and Oh, H.C. (1993) Do American business travellers have different hotel service require-

ments? International Journal of Contemporary Hospitality Management 5, 16–21.Wind, J., Green, P.E., Shifflet, D. and Scarbrough, M. (1989) Courtyard by Marriott: designing a hotel facil-

ity with consumer-based marketing models. Interfaces 19, 25–47.Woodside, A.G. and Lysonski, S. (1989) A general model of traveler destination choice. Journal of Travel

Research 27, 8–14.



Chapter threeThe Influence of Consumers’ Emotions on their ServiceProduct Evaluation

Sandra Gountas1 and John Y. Gountas2

1Department of Marketing, Monash University, Melbourne, Victoria, Australia; 2BowaterSchool, Deakin University, Geelong, Victoria, Australia

Abstract

There is a lack of research on the role of emotions in consumers’ product evaluation. Much of the servicesevaluation literature focuses on the tangible and intangible aspects of the services and their effects on con-sumer satisfaction. However, these effects are contingent upon a wide range of factors, including the dura-tion of the service, the individual’s personality, natural preferences and their emotional states prior to,during and after the service encounter. The leisure airline industry provides an ideal setting to study theeffects of the aforementioned factors. This chapter reports the findings of a study that currently comprisesmore than 1400 cases and focuses on the influence of emotions on service evaluation. The findings indi-cate that leisure airline passengers’ emotions prior to and during the flight are related to the service provi-sion, and that emotions are related to the passengers’ overall satisfaction rating for the services received.

Introduction: Emotions and Services Satisfaction

Many authors have noted that more researchis needed into the effects of emotions and/ormood state on consumer behaviour(Gardner, 1985; Dube and Morgan, 1998,Fournier and Mick, 1999; Holbrook andGardner, 2000; Howard and Gengler, 2001).In her paper on mood states and consumerbehaviour, Gardner (1985, p. 281) remarked,‘Individuals often try to anticipate each oth-ers’ moods prior to interactions and to readeach others moods during encounters.’ Thisis something that most people instinctively

understand from their personal interactions.However, this dynamic is not always consid-ered in service encounters research. Gardner(1985) goes on to point out that mood-stateknowledge may be salient to understandingconsumer behaviour. Service encounters arecomplex and evoke a range of emotionalresponses as well as cognitive evaluation(Derbaix and Pham, 1991; Rust et al., 1995;Price et al., 1999).

Clearly, it would be useful if a serviceprovider understands the level of impact thata particular mood state has on the consumers’service evaluations. In the case of an airline, itwould be useful to understand the influence


Consumer Psych - Chap 03 5/12/03 11:53 am Page 21

of positive or negative emotions on the con-sumer, and to develop strategies to anticipateand respond to mood states as part of the nor-mal service process and in service recovery.

The consumption of many services is oftenconnected with leisure or discretionaryincome. In this case, the consumption act maybe more highly charged with affective conno-tation than in the consumption of non-discre-tionary products. Price et al. (1999) describedthe service encounter provided by flight atten-dants (airlines) as a low-affect, social/publicdistance extended duration service. However,leisure airline services are high-affect serviceencounters. This is due to the proportionallyhigh level of discretionary income expendi-ture, the frequently lengthy anticipation ofthe service experience and the individual’sexpectation of the product benefits.

Mehrabian and Russell’s (1974) work onenvironmental psychology identified threevariables that describe affective response toenvironments. These are pleasure, arousaland dominance (PAD). Pleasure is describedand measured by expressions such as happy,pleased and satisfied. Arousal is described bythe extent to which the consumer is stimu-lated, e.g. fear, curiosity, etc. Dominance isconcerned with the individual’s feeling ofcontrol over the situation being experienced.As an example, air travel is likely to be a situa-tion over which a consumer feels very little orno control, and which may evoke a range ofemotions and mood state responses.

Clearly, in the case of a leisure flight (alsoknown as a charter flight in some regions)consumers could easily be happy but a littlefearful (of flying), while feeling moderatelyaware of their lack of control over the situa-tion. On the other hand, consumers mayhave experienced a long delay before depar-ture, leaving them displeased, anxious andacutely aware of their lack of control over thesituation. The different combinations andintensities of emotions may affect both theconsumer’s behaviour during product con-sumption and his or her overall satisfaction.

The duration of leisure or charter flightsmay range from less than an hour to 23 hfrom one side of the world to the other. Theair travel experience includes the actualflight, as well as check-in at the airport, wait-

ing in the airport to board the aircraft, etc.This allows time for changes in moodstate/emotions to develop. Holbrook andGardner (2000) distinguished between moodand emotion by referring to mood as con-stantly evolving general affective states felt byindividuals and to emotion as specific affec-tive responses prompted by particular con-sumption experiences. This methoddifferentiates between a milder, more diffuse-feeling state (mood) that may influenceand/or reflect what is happening around theconsumer and a more intense, object-specific-feeling state (emotion) that responds to par-ticular consumption activities. This wouldseem to suggest that one (mood) might be aprecursor to the other (emotion).

When using the example of leisure airtravel, it is probably also useful to considerthe issue of emotional contagion or emo-tional infection. Travelling in an aircraft withother people in close proximity is likely to fos-ter some spreading of positive or negativeemotions. Howard and Gengler (2001) exam-ined the relationships between ‘receivers’ and‘senders’ of a happy emotion. This notionmay be considered in terms of the relation-ships between the service provider and con-sumers, as well as the relationships betweenthe consumers themselves. Furthermore,Neumann and Strack (2000) have high-lighted the differences between expectationsfor mood versus emotional contagion.

There has been much written about theimpact of expectation and anticipation onleisure travellers (Shiv and Huber, 2000).Emotional contagion may be induced by oth-ers known to, or with, the consumer at thetime of consumption or by total strangers in adensely populated environment such as anaircraft. Arnould and Price (1993) consid-ered that affective interactions were useful infully understanding satisfaction with a serviceexperience. These interactions may bebetween service provider and consumer, pos-sibly also including the interactions withother consumers.

Holbrook and Gardner (2000, p. 166) ask,‘If moods are output variables that character-ize important affective consequences of con-sumer behaviour, how do these moodschange in response to the progression of con-

22 S. Gountas and J.Y. Gountas


sumption experiences? If the consumerbegins in a particular mood, what accountsfor alterations in this mood? By what processdo moods develop over time?’ The mostimportant question is ‘How do these moods,emotions, and processes affect the con-sumer’s level of satisfaction?’.

Plutchik (1980) and Russell (1978) con-nected satisfaction and emotional state, withsatisfaction being more specific in connota-tion than dissatisfaction. This suggests thatdissatisfaction and its antecedents are morecomplex than the factors that result in satis-faction. Plutchik’s scale of emotions is mea-sured using the following triads of objectives(Sheth et al., 1999):

Fear: threatened, frightened, intimidatedAnger: hostile, annoyed, irritatedJoy: happy, cheerful, and delightedSadness: gloomy, sad, and depressedAcceptance: helped, accepted, and trustingDisgust: disgusted, offended, unpleasantAnticipation: alert, attentive, curiousSurprise: puzzled, confused, startled

Izzard (1977) developed the DifferentialEmotional Scale (DES II) which consists often emotions which incorporate Plutchik’seight primary emotions of Fear, Anger, Joy,Sadness, Acceptance, Disgust, Anticipationand Surprise but also include Contempt,Guilt and Shame as discrete emotions. Whatis interesting is that neither researcher identi-fied the opposite emotion to Disgust, which isDignity: self-respect, pride and self-assurance.Many consumers are concerned about beingvalued and respected by being acknowledgedand appreciated as valued customers.

Havlena and Holbrook’s (1986) study,comparing Mehrabian and Russell’s andPlutchik’s scales, used 20 individuals to pro-duce descriptions of their consumption expe-riences in a number of contexts includingchurch, athletics, entertainment, dining, hob-bies, fashion and security. A total of 149 ‘realdescriptions’ were produced and analysed bytwo groups of ten judges drawn from MBAstudents at a major American university, con-cluding that Mehrabian and Russell’s frame-work is more robust than Plutchik’s. However,as pointed out by Hair et al. (1995), the use ofstudent samples limit the generalization ofthis conclusion.

The following propositions were devel-oped from the literature review:

P1 Positive and negative emotions arerelated, respectively to positive and neg-ative service evaluation.

P2 Emotional states are antecedents of lev-els of reported satisfaction.

P3 Emotional contagion influences othercustomers’ emotional states duringextended-duration service.

Methodology

The first stage of this study included secondaryresearch on published material and an inter-nal database from previous customer satisfac-tion questionnaires from a major leisureairline in the UK. The second stage involvedprimary qualitative research using focusgroups from two major market segments in thesouth and north of England. Two sets of focusgroups were conducted with past customersfrom the leisure airline’s database. The selec-tion of all the participants was via stratifiedrandom probability methods. The first set ofeight focus groups took place during June2000 in London and Manchester, UK. Eachfocus group comprised between eight and tenparticipants. The second round of the qualita-tive research comprised four focus groups witheight to ten participants in each group, andtook place during February 2001 in Londonand Manchester. The participants wereselected from the airline’s database of past pas-sengers and represented the airline’s ‘typical’passengers in gender, age, party composition,occupation (social group) and service experi-enced in the range of all flight sectors and ser-vice levels. A semi-structured research guidewas used to examine/identify the affective,tangible and temporal aspects of the serviceencounter and the evaluation criteria appliedby past consumers.

Qualitative research findings from the focusgroups’ discussions

Most of the participants in the 12 focusgroups (n=110) mentioned similar issuesabout, and criteria for, service evaluation.The findings are grouped in Table 3.1. The

Influence of Consumers’ Emotions on Satisfaction 23


emotions and mood states of the passengersappeared to be very important in their eval-uation of the overall service, as did the tem-poral aspects. Longer flights, take-off delaysand waiting at the check-in desk seemed totrigger mood responses that affected overallsatisfaction.

Travelling companions and other passen-gers also seem to influence the individual’smood, emotions and attitude towards theflight attendants and their service provision.The focus groups indicated interactionsbetween service levels received, duration andother people’s moods and behaviours mayresult in mood or emotional changes. In com-bination with the literature review, the focusgroup findings led to the development of thefield research propositions.

The survey instrument

As with Liljander and Strandvik’s study(1997), the SERVQUAL dimensions ofReliability, Responsiveness, Assurance,Empathy and Tangibles have been used, inthe construction of the new questionnaire.The items within these dimensions partly

express concerns repeatedly raised duringthe focus group conducted by the researcher.However, these dimensions do not addressfully the other aspects of the seven categoriesidentified by the focus group research. Inorder to ameliorate the deficit, the question-naire was developed to measure service qual-ity, expectations, the existence and impact ofemotions during service consumption, indi-vidual personality traits and other psychologi-cal factors derived from perceptions of pastand present service experiences, the serviceprovider’s image and personal space needs.

Plutchik’s (1980) emotions scale was usedto measure the impact of emotions andchanges in emotions over a period of time.This scale was chosen because of the similar-ity of expression used by Plutchik with that ofthe focus group participants. The adjectivesfrightened, irritated, happy, sad, trusting, offended,curious and confused, from Plutchik’s scale,were used frequently by the focus participantsto express their emotions or mood state andwere thus deemed appropriate for use in thequantitative survey. Furthermore, unlikeMehrabian and Russell’s scale (Havlena andHolbrook, 1986), Plutchik’s scale includesAnticipation or Expectancy, which has been


Table 3.1. Service criteria.

Hygiene factors Service/empathy Psychological factors

Aircraft Personalization Perception of airlineCabin appearance Assurance Previous experienceCrew appearance Responsiveness ExpectationsFood Service level ConfidenceEntertainment Personality typesIn-flight magazine ToleranceChildren’s amenities CrowdingOn-board shopping Emotions/mood stateIn-flight magazine

Customer’s personalcharacteristics Situation specifics Temporal factors

Age Party composition Check-in durationGender Flight sector Flight delaysMarital status Purpose of travel Flight durationIncome Other passengers In- or out-bound flightOccupation Service delaysNationality Changes to expected servicePhysical size Weather conditions


considered an important influence on aleisure traveller’s or tourist’s evaluation of aservice (Tribe and Snaith, 1998; Weiermairand Fuchs, 1999).

Although Izzard’s (1977) DES II has beenfound to be robust for measuring emotionsduring consumption, the focus group partici-pants did not refer to these emotions(Contempt, Guilt and Shame); as such, theirinclusion was considered inappropriate forthis particular setting.

The structure of the airline flight satisfactionquestionnaire

The questionnaire consisted of five mainsections:

● The first section had 28 positively wordeditems concerned with the participants’perception of the airline, empathy withthe crew, general attitudes towards flyingas a form of travel and feelings of confi-dence in the product.

● The second section was concerned withthe participants’ expectations and the fac-tors that may have contributed towardsthose expectations.

● The third section was designed to measurethe differences in emotions before andduring the flight, and identify the causesof change. A question on overall satisfac-tion was included.

● The fourth section consisted of personal-ity-type orientation statements.

● The fifth section was concerned with thedemographic profile of the participants.

Item measurement scales

Parasuraman (1995) examined service qualityas a function of the expectation–performancegap. SERVQUAL was subsequently designed tomeasure the difference between expectationand performance. There is debate about theusefulness and validity of this method(Carman, 1990; Cronin and Taylor, 1992;Peter et al., 1993 etc.). Cronin and Taylor(1992) asserted that their performance-basedscale (SERVPERF) was more efficient than thedifference-based scale (SERVQUAL). For the

purpose of this research, a compromisebetween measuring performance only andmeasuring the difference between expecta-tions and perceived service was deemedappropriate. The reasons for this decision are:

● Exploring how expectations are formedand whether they affect perceived servicequality and satisfaction is undeniably use-ful to the service provider.

● The research hypotheses require a mix-ture of performance assessment and gapanalysis to be confirmed or disconfirmed.

● The practical aspects of the empiricalresearch for data collection lend them-selves more easily to a mix of performanceand gap analysis.

The items are scored using a ten-pointLikert scale with anchors of strongly disagreeand strongly agree. The ten-point scale waschosen for its ease of use and because ten-point scales have previously been shown tohave a high predictive and convergent validity(Parasuraman, 1995). The ten-point scaleallows the participants to express their evalua-tion with a greater degree of differentiation.

Data Collection Method

The sample was from the same UK leisure air-line’s customer base used for the exploratoryfocus groups. In total, 5000 questionnaireswere distributed on a range of short- andlong-haul leisure or charter flights to differentdestinations mostly departing from the UKduring the months of October, November andDecember 2001. The sampling method usedwas a probability, stratified random methoddrawing from all the flights available from allthe UK regional airports and including allflight sectors for short- and long-haul destina-tions. This method allows full representationof all geographic and demographic character-istics of the UK leisure flight market.

Of the 5000 questionnaires distributedduring the outbound flights by the desig-nated crew members, 1773 were returned, areturn rate of 35%. However, 346 of thesequestionnaires were deemed unusable eitherbecause of missing values or due to therespondent being under 16 years of age. After



the data cleaning process, there were 1427usable questionnaires, i.e. 28% of the numberdistributed.

Generally, the quality of responses wasvery comprehensive. However, a small num-ber of respondents had left a few questionsunanswered. For example, according to therespondents’ comments, a missing explana-tion/response was often due to the lack ofpersonal experience; therefore, no com-ment/answer was provided. More specifically,for question 6, ‘Air 2000 staff members havethe power to solve problems’ – if the respon-dent had not encountered any problems, theyfelt unable to answer the question. Similarly,for question 29, ‘My expectations of the air-line’s service are influenced by:…’, therespondents sometimes rated only theoptions that applied to them, leaving the oth-ers blank. This is not considered a problem asthe options can easily be isolated for the pur-pose of analysis; alternatively, it is possible toinfer a value from the values on other vari-ables (Sapsford and Jupp, 1996).

It is reasonable to assume that if respon-dents gave a value of 10 to their expectationsbeing influenced by brochures, a value ofeight to TV programmes and no value to theother options, that the other options werenot perceived to be a significant influence ontheir expectations of the airline’s service,which should not bias the conclusions in anyway. Similarly, this situation applies to ques-tions 30a/d, ‘This flight, today has met myexpectations of…,’; question 32a/h; ‘Justbefore I got on this flight, I felt:…’; question33a/h, ‘I now feel (at this moment):…’; ques-tion 34a/h, ‘My feelings have changedbecause of:…’; and the personality items inquestions 37 and 38. It is assumed that, inmany cases, the respondents elected toanswer only the items that are important tothem, due to the consistency of theirresponse pattern.

Research Findings

In order to identify the salient patternsrelated to the emotional issues and servicesatisfaction, the pre- and during flight emo-tions were factor analysed. Oblique factor

rotation was used, as it is more flexible thanorthogonal methods. The oblique method,Equimax, is more realistic because it deter-mines the factors according to the underpin-ning theoretical constructs, assuminginter-correlations between the factors, whichis appropriate for this case. The factors pro-duced using Equimax should allow meaning-ful interpretation (Hair et al., 1995).

Four factors emerged and were groupedby negative and positive emotional states(Table 3.2). Factor 1 relates to theDispleasure and Confusion emotions andproduced lower values for the pre-flight andhigher for during the flight service experi-ences. For example, the factor 1 loadingscores for the question ‘displeasure/offendedbefore the flight’ is 0.66, as opposed to thefeelings of ‘displeasure/offended’ during theflight, which is higher, 0.80. Likewise, thestatement ‘Before the flight I felt irritated’with a factor loading of 0.53 and during theflight experience the score for ‘I now feel irri-tated’ is higher at 0.69.

In factor 2, regarding Sadness and Fear,the values of the statements just before theflight tend to be higher than the valuesexpressed during the flight service experi-ence. For example, the statement ‘Just beforeI got on this flight I felt sad’ produced a fac-tor loading score of 0.73, as opposed to thestatement of ‘I now feel sad’ with a lowerscore of 0.60.

Factor 3, questions related to Happinessand Trust, followed a similar pattern to factor1. The pre-flight factor loading score ishigher at 0.83 for ‘Before I got on this flight Ifelt happy’, as opposed to the lower score of0.72, for ‘I now feel happy.’

Factor 4, dealing with the issues ofCuriosity, produced identical factor loadingscores of 0.83, for the emotions/feelings ofcuriosity before and during the flight.

The correlations of the four factors relatedto the emotional states and overall satisfac-tion produced some interesting findings.Table 3.3 indicates that the strongest correla-tion with overall satisfaction is with the posi-tive emotions of felt happiness and trust. Thehighest negative correlation is with the nega-tive emotions of displeasure/offended, irri-tated and puzzled/confused.



Two factors emerged concerning reasonsfor reported emotional change. These areshown in Table 3.4.

Tables 3.5 and 3.6 show the correlationsbetween the pre- and during consumptionemotions and overall satisfaction with the ser-vice. There is a positive correlation betweenpositive emotions and overall satisfaction, anda negative correlation between negative emo-tions and satisfaction. It is interesting to notethat trust is the only emotion that remainedunchanged for pre- and during consumptionstages. Curiosity decreased slightly during

consumption but the remaining emotionsincreased in strength during the flight experi-ence stages.

In order to identify whether there are anysignificant relationships between the emo-tions and the reasons for the changes and theoverall perception of satisfaction, a correla-tion analysis was carried out (Table 3.7).

Overall, the most pronounced reasons forinfluencing satisfaction are ‘my feelingschanged because of the in-flight service andcheck-in services, my travelling companion’spositive mood and being close to other people’.


Table 3.2. Factor analysis of all the emotions before and during the flight serviceexperience.

Factor loadings

Factor 1: Displeasure and ConfusionJust before I got on this flight I felt displeased/offended 0.66I now feel displeased/offended 0.80Just before I got on this flight I felt irritated 0.53I now feel irritated 0.69Just before I got on this flight I felt puzzled/confused 0.54I now feel puzzled/confused 0.60

Coeff. alpha 0.84Variance explained (%) 19.1Eigenvalue 3.05

Factor 2: Sadness and FearJust before I got on this flight I felt sad 0.73I now feel sad 0.60Just before I got on this flight I felt frightened 0.72I now feel frightened 0.71


Factor 3: Happiness and TrustJust before I got on this flight I felt happy 0.63I now feel happy 0.72Just before I got on this flight I felt trusting 0.84I now feel trusting 0.83


Factor 4: CuriosityJust before I got on this flight I felt curious 0.83I now feel curious 0.83


Extraction method: principal component analysis; rotation method: Equimax withKaiser Normalization.



Table 3.3. Correlations between the four emotions factors and overall satisfaction.

Pearson correlation Sig. (two-tailed)

Factor 1 (Displeasure and Confusion) and overall satisfaction �0.355 0.000Factor 2 (Sadness and Fear) and overall satisfaction �0.010 0.766Factor 3 (Happiness and Trust) and overall satisfaction 0.515 0.000Factor 4 (Curiosity) 0.036 0.266

Table 3.4. Factor analysis of the reasons for reported emotional changes.

Factor 1: Intangible service characteristics and other people’s influenceMy feelings have changed because of the check-in service 0.86My feelings have changed because of the in-flight service 0.85My feelings have changed because of my companion’s positive mood 0.82My feelings have changed because of being very close to other people 0.63


Factor 2: Other passengers behaving badly and boredomMy feelings have changed because of other people behaving badly 0.74My feelings have changed because of my becoming bored during the flight 0.79My feelings have changed because of other reasons 0.75


Table 3.5. Correlations: reported emotional state prior to consumption and overall satisfaction.


Just before I got on this flight I felt happy 0.385 0.000Just before I got on this flight I felt irritated �0.188 0.000Just before I got on this flight I felt curious 0.027 0.380Just before I got on this flight I felt frightened �0.003 0.923Just before I got on this flight I felt trusting 0.424 0.000Just before I got on this flight I felt displeased/offended �0.222 0.000Just before I got on this flight I felt puzzled/confused �0.158 0.000Just before I got on this flight I felt sad �0.138 0.000

Table 3.6. Correlations: reported emotional state during the flight and overall satisfaction.


I now feel happy 0.596 0.000I now feel irritated �0.397 0.000I now feel curious 0.005 0.875I now feel frightened �0.103 0.001I now feel trusting 0.424 0.000I now feel displeased/offended �0.405 0.000I now feel puzzled/confused �0.209 0.000I now feel sad �0.251 0.000


Discussion

Proposition 1: ‘Emotions are related to positiveand negative evaluations of services’, is sup-ported partially. The negative emotionalstates of feeling displeased and confused indi-cated a statistically significant negative corre-lation of r = �0.355. The positive emotionsof Happiness and Trust produced a strongpositive correlation of r = 0.515, indicatingthat they are factors influencing the overallsatisfaction. However, the other two factors ofSadness and Fear, and Curiosity did not pro-duce any significant correlations.

It appears that only the strongest emo-tional states of Displeasure, Confusion,Happiness and Trust are critical influences inthe overall perception of satisfaction. Thismay be due to the consumer’s attribution andperception of equity with the service transac-tion since it could be argued that whilst dis-pleasure, confusion, happiness and trust arelikely to be in the sphere of the serviceprovider’s influence the other emotions mayexist independently.

Proposition 2: ‘Emotional states areantecedents of levels of reported satisfaction’.This is supported partially by the findings onthe emotional states before and during theflight services and overall satisfaction. Theemotions prior to the flight with the highestpositive correlations are: ‘I felt happy’, r =0.385, and ‘I felt trusting’, r = 0.424. The sig-nificant negative correlations that influencesatisfaction are: ‘I felt displeased/offended’,r = �0.222; ‘I felt irritated’ r = �0.188; ‘I feltpuzzled/confused’, r = �0.158; and ‘I felt

sad’, r = �0.138. The effect of emotions dur-ing the service delivery/interaction is greaterin significance with service delivery, accord-ing to the correlations on Table 3.6. The posi-tive emotions of happiness is r = 0.596 andtrusting is r = 0.424 (the same as the pre-flight score). The increase in the correlationbetween feeling happy and satisfaction beforeand during the flight is nearly double.

The negative emotional states alsoincreased during the flight. The feeling ofbeing irritated has a correlation of r =�0.397; ‘I now feel displeased/offended’, r =�0.405; ‘I now feel puzzled/confused’, r =�0.209; and ‘I now feel sad’ is r = �0.251.

The impact of the service experience onconsumers’ emotions and satisfaction evalua-tion is clear and indicates real potential forintervention and recovery practices havingpositive outcome.

The most important priority is the on-board service quality provision, but the pre-flight service experiences are also significantin the overall perception of satisfaction.Customers need a ‘seamless’, efficient servicefor all stages of the service transactions. Eachstage of the service transaction has a degreeof influence and the overall cumulative expe-rience influences the overall satisfaction level.

Proposition 3: ‘Emotional contagion influencesother customers’ emotional states duringextended duration service’. This is supportedby the findings shown in Table 3.7. The pre-and during flight services are highly corre-lated with the influences of travelling com-panions’ positive mood, r = 0.612 and r = 0.596respectively. The influence of the travelling


Table 3.7. Correlations between the reasons for emotional change and overall satisfaction.


My feelings have changed because of the in-flight service and 0.776 0.000because of the check-in service

My feelings changed because of the in-flight service and 0.612 0.000because of my travelling companion’s positive mood

My feelings have changed because of the check-in service and 0.596 0.000because of my travelling companion’s positive mood

My feelings have changed because of being very close to other people 0.503 0.000and because of my travelling companion’s positive mood

My feelings changed because of the in-flight services and 0.409 0.000 because of being very close to other people


companions’ positive mood and being closeto other people produced a correlation of r =0.503. The in-flight service experience andthe feeling of being close to other peopleproduced a correlation of r = 0.409. All corre-lations are significant with high scores. Thisindicates that the social aspects/factors aresignificant and, therefore, the service man-agers need to think about the social needsand influences of others as well as the passen-gers as individuals.

Further analysis of the data shows thatthe intangible aspects of service, i.e. the‘people’ who are delivering the service, aremore important to overall satisfaction thanthe tangible aspects such as the physicalenvironment. These findings also confirmthe assumptions made at the exploratorystage of research.

The implication of the results for practi-tioners is that the aspects of service that influ-ence emotions prior to, and during,consumption need to be carefully identifiedand monitored, as emotions have a significantimpact on consumers’ satisfaction levels. Thismeans that the traditional comparison stan-dards paradigm may be inadequate as a singlemethod for understanding what contributesto and influences satisfaction (Fournier andMick, 1999).

It is necessary to consider industry-spe-cific issues. In the case of leisure airlines,competition is intense. There are many com-petitors offering almost identical products.The airlines are often part of vertically andhorizontally integrated companies, whichare complex in structure, placing constraintson the scope of product development anddelivery.

These issues are exacerbated by the tradi-tional pricing strategy of this industry, whichleaves the airlines operating on very lowprofit margins. Most airlines are dependenton economies of scale. The extended dura-tion nature of the total service from decisionmaking through to purchase through to con-sumption means that there are numerouspoints in the process at which an extraneouselement may interfere with the airline’s abil-ity to influence the consumer’s overall evalua-tion.

In summary, the study has shown thatthere are very strong relationships betweenthe consumers’ emotions and perceived satis-faction, and that these vary with differentaspects of services and are, in some part,influenced by other consumers as well as ser-vice providers. The next stage of the researchwill attempt to measure the extent of thisinfluence more precisely.


References

Arnould, E. and Price, L. (1993) River magic: extraordinary experience and the extended serviceencounter. Journal of Consumer Research 20, 24–45.

Carman, J.M. (1990) Consumer perceptions of service quality: an assessment of the SERVQUAL dimen-sions. Journal of Retailing 66, 33–55.

Cronin, J.J. and Taylor, S.A. (1992) Measuring service quality: a re-examination and extension. Journal ofMarketing 56, 55–68.

Derbaix, C. and Pham, M.T. (1991) Affective reactions to consumption situations: a pilot investigation.Journal of Economic Psychology 12, 325–355.

Dube, L. and Morgan, M.S. (1998) Capturing the dynamics of in-process consumption emotions and satis-faction in extended service transactions. International Journal of Research in Marketing 15, 309–320.

Fournier, S. and Mick, D.G. (1999) Rediscovering satisfaction. Journal of Marketing 63, 5–23.Gardner, M.P. (1985) Mood states and consumer behaviour. Journal of Consumer Research 12, 281–300.Hair, J.F., Anderson, R.E., Tatham, R.T. and Black, W.C. (1995) Multivariate Data Analysis with Readings, 4th

edn. Prentice Hall, Sydney. Havlena, W.J. and Holbrook, M.B. (1986) The varieties of consumption experience: comparing two typolo-

gies of emotion in consumer behaviour. Journal of Consumer Research 13, 394–404.Holbrook, B. and Gardner, M.P. (2000) Illustrating a dynamic model of the mood-updating process in con-

sumer behaviour. Psychology and Marketing 17, 165–194.


Howard, D.J. and Gengler, C. (2001) Emotional contagion effects on product attitudes. Journal of ConsumerResearch 28, 189–211.

Izzard, C.E. (1977) Human Emotions. Plenum Press, New York.Liljander, V. and Strandvik, T. (1997) Emotions in service satisfaction. International Journal of Service

Industry Management 8, 148–169.Mehrabian, A. and Russell, J.A. (1974) An Approach to Environmental Psychology. MIT, Cambridge,

Massachusetts.Neumann, R. and Strack, F. (2000) Mood contagion: the automatic transfer of mood between persons.

Journal of Personality and Social Psychology 79, 211–235.Parasuraman, A. (1995) Measuring and monitoring service quality. In: Glynn, W.J. and Barnes, J.G. (eds)

Understanding Services Management. John Wiley & Sons, Chichester, pp. 143�177.Peter, J.P., Churchill, G.A. and Brown, T.J. (1993) Caution in the use of difference scores in consumer

research. Journal of Consumer Research 19, 655–662.Plutchik, R. (1980) Emotion: a Psychoevolutionary Synthesis. Harper & Row, New York.Price, L.L., Arnould, E.J. and Tierney, P. (1999) Going to extremes: managing service encounters and

assessing provider performance. In: Bateson, J. and Hoffman, D. (eds) Managing Services Marketing.Dryden Press, Orlando, Florida, pp. 249–266.

Russell, J.A. (1978) Evidence of convergent validity on the dimensions of affect. Journal of Personality andSocial Psychology 36, 1152–1168.

Rust, R.T., Zaborik, A.J. and Keinningham, T.I. (1995) Return on quality (ROQ): making service qualityfinancially accountable. Journal of Marketing 59, 58–70.

Sapsford, R. and Jupp, V. (1996) Data Collection and Analysis. Sage Publications, London.Sheth, J.N., Mittal, B. and Newman, B.I. (1999) Consumer Behaviour: Consumer Behaviour and Beyond. Dryden

Press, Orlando, Florida.Shiv, B. and Huber, J. (2000) The impact of anticipating satisfaction on consumer choice. Journal of

Consumer Research 27, 202–223.Tribe, J. and Snaith, T. (1998) From SERVQUAL to HOLSAT: holiday satisfaction in Varadero, Cuba.

Tourism Management 19, 25–34.Weiermair, K. and Fuchs, M. (1999) Measuring tourist judgment on service quality. Annals of Tourism

Research, 26, 1004–1021.



Chapter fourValidating a Guttman-type Social Distance Scale forExplaining Residents’ Attitudes towards Tourism

Maree Thyne1 and Andreas H. Zins2

1Scottish Centre of Tourism, The Robert Gordon University, Garthdee II, Garthdee Road,Aberdeen AB10 7QG, UK; 2Institute for Tourism and Leisure Studies, University of Economicsand Business Administration, Augusse 2–6, A-1090 Vienna, Austria

Abstract

Research explaining host-community residents’ attitudes towards tourism receives continuous attentionand its scope is broadening constantly. This study aims at contributing to the body of knowledge about thesystematic variation and change of these attitudes. For this reason, the concept of social distance is intro-duced into the tourism context. As a first attempt, a Guttman-type scale with Thurstone characteristics wasdeveloped and purified in two countries: New Zealand and Austria. In both countries, independent sur-veys were conducted to validate the scale formation. Results with regard to the scale properties are encour-aging, though improvements and more cross-validation has to be done. The contribution to explainingresidents’ attitudes was very fruitful in the Austrian but less clear in the New Zealand example. Whetherdifferent stages of tourism development are responsible for these differences cannot be entirely concludedfrom only two surveys.

Introduction

The sustainable tourism literature has pro-posed and tested the effects of a broad scopeof factors on residents’ opinions and attitudestowards tourism. Individual characteristics,community features, as well as factors cover-ing relationships between the residents andthe community, can be responsible for per-ceiving more or less negative and positiveimpacts from tourism activities. It is the aimof this chapter to investigate whether another

personal factor has the strength to contributeto this explanation: social distance.

Although social distance is not a new issuein the social sciences, it entails an innovationwithin tourism literature. This chapterfocuses on the theoretical anchors of the con-cept and the careful steps of scale develop-ment. The technical concept followsGuttman’s and Thurstone’s suggestions,resulting in a uni-dimensional, hierarchicaland cumulative scale with almost equal inter-vals. The procedure and outcome of the



development of a social distance measure-ment instrument are compared between twocountries: New Zealand and Austria. Externalvalidity was investigated by correlationalanalyses with the perceived impacts fromtourism and attitudes towards future tourismdevelopment.

Theoretical Concepts

Social distance

Over the past 65 years, the concept of socialdistance has been adapted to aid in the mea-surement of cross-cultural impacts. Bogardus(1940) defines social distance as a conceptthat explains the cooperative behaviour andsympathetic understanding that existsbetween people. Although the concept hasbeen used in a variety of contexts, includingsocial distance towards rural life (Binnewies,1926, cited in Anderson, 1983), social dis-tance in the Shakespeare play Much Ado aboutNothing (Bogardus, 1934) and social distancebetween children and their parents (DuVall,1937, cited in Anderson, 1983), it has notbeen extensively applied in tourism settings.The most common context in which socialdistance has been employed is the study ofrace, nationality and cross-cultural contact(Bogardus, 1925; Crull and Bruton, 1979;Owen et al., 1981).

The predominant reason for using socialdistance scales in determining attitudestowards cross-cultural contact are first,because it has often been advocated thatsocial distance encompasses and is easier tomeasure than such concepts as prejudice,racism, discrimination, stereotyping and eth-nocentrism (Triandis and Triandis, 1960;Jaspars and Hewstone, 1982; Dyer et al., 1989;Hraba et al., 1989, 1996; Hagendoorn andKleinpenning, 1991; Lemaine and Ben Brika,1997). Second, authors have stated that socialdistance scales are simple scaling devices, eas-ily understood by a wide range of people, andnoted for consistent results and reliability(e.g. Anderson, 1983; Hagendoorn andKleinpenning, 1991).

The term ‘social distance’ was first coinedby Robert Park (1924, p. 340) who believed

that we not only ‘have a sense of distancetoward individuals with whom we come intocontact, but we have much the same feelingwith regard to classes and races’. EmoryBogardus operationalized this concept in thefollowing year and began his research onsocial distance: research that would cover a40-year period (Bogardus, 1967). His workwas inspired by a concern with what hereferred to as ‘the race problem’, which hebelieved to be one of the major social dilem-mas confronting America at the time(Bogardus, 1922).

In 1925, Bogardus developed his firstsocial distance scale of which several differentversions were tried over the following decade.Although adaptations of Bogardus’ scale,which was finally administered in 1933, wereadopted by a number of authors over the fol-lowing 50 years, none were specific enough tobe adopted for the context of this chapter.

Unlike Bogardus’s first scale (seeBogardus, 1933, for a discussion of hismethodology), the majority of subsequentsocial distance scales have not been throughrigorous development processes; rather,scales have often been borrowed from previ-ous research and put into very different con-texts without appropriate alteration (e.g.Prothro and Miles, 1953; Vaughan, 1962;Owen et al., 1981). Additionally, manyresearchers give very few details on how theydesigned their social distance scale (e.g.Trlin, 1970; Dyer et al., 1989; Hraba et al.,1989; Nix, 1993). More recently however,there has been greater emphasis placed onscale reliability (e.g. Jacoby, 1978; Churchill,1979; Lankford and Howard, 1994; Ap andCrompton, 1998; Williams and Lawson,2001). Given the new context of this socialdistance study (tourism), a rigorous develop-ment process was adopted. The qualitativeand quantitative phases employed in thatprocess are the subjects of this chapter.

Attitudes toward tourism

Research on the phenomena induced bytourism activities and development withinhost communities has a tradition of morethan 25 years. This study does not focus on

34 M. Thyne and A.H. Zins


the evolution and differentiation of the resi-dent’s view or attitudes toward tourism; itsolely refers to the framework for monitoringcommunity impacts of tourism within whichthe concept of social distance towards touristsmay contribute to explain the variation inattitudes. Hence, it may suffice here to iden-tify some literature offering an overview ofthe topic complemented by some recent pub-lications (Table 4.1).

In studying perceived residents’ impactsinduced by tourism activities, a variety ofissues has been raised. Whereas some empha-sis was laid on developing a standardizedmeasurement instrument (Lankford andHoward, 1994; Rollins, 1997; Schneider et al.,1997), numerous empirical studies use differ-ent sets of statements covering a differentscope of impacts. However, convergenceseems to evolve upon the basic dimensionsunderlying this particular attitudinal space.Factors with overlapping and, therefore, cor-relating content are frequently reported.

The predominant aspect covers the eco-nomic impacts of tourism. They can beattached to collective consequences or toindividual benefits (e.g. income, job opportu-nities) or disadvantages (e.g. rising livingcosts), being either monetary or non-mone-tary in nature. A somewhat related matter

addresses the changes in infrastructure (e.g.improved roads and other public services,sports and leisure facilities).

The environmental dimension has broad-ened its scope over the years as well: frompollution and natural deterioration, theissues shifted towards traffic and othercrowding problems. The last – and mostrecent – area of interest focuses on the socialand cultural impacts from tourism. Therange of topics goes from changes in com-munity life and social roles, to discriminationbetween tourists and residents, and effectson cultural aspects such as norms, crime andarchitecture.

Perceived impacts of tourism can be seenas one perspective on how attitudes towardtourism are captured. Most of the proposedmeasurement instruments care for the per-ceived degree of various outcomes; however,most did not explicitly incorporate the sub-jective valuation of these outcomes. Anotherperspective considers attitudes towardtourism development, asking residentswhether and how they prefer future tourismactivities. Whereas the classical TIAS instru-ment (Lankford and Howard, 1994) mixesboth aspects, other studies try to addressthese kinds of attitudes separately (Perdue etal., 1990; Getz, 1994; Lindberg and Johnson,

Social Distance in Attitudes towards Tourism 35

Table 4.1. Dimensions and main issues of perceived impacts of tourism.

Dimension Literature

Positive and negative economic impacts Ap and Crompton, 1998 (literature summary); Lindbergand Johnson, 1997; Bachleitner and Zins, 1999

Community, region Faulkner and Tideswell, 1997; individual Lawson et al., 1998Employment Ap and Crompton, 1998; Faulkner and Tidewell, 1997;

Lawson et al., 1998Positive and negative environmental impacts Ap and Crompton, 1998 (literature summary);

Bachleitner and Zins, 1999; Yoon et al., 1999CrowdingCongestion Lindberg and Johnson, 1997Pollution Faulkner and Tideswell, 1997

Positive and negative social and cultural impacts Ap and Crompton, 1998 (literature summary);Bachleitner and Zins, 1999; Yoon et al., 1999

Crime Lawson et al., 1998Discrimination Bachleitner and Zins, 1999Aesthetics Lindberg and Johnson, 1997

Infrastructure and community services Bachleitner and Zins, 1999; Faulkner and Tideswell,1997; Lawson et al., 1998; Yoon et al., 1999


1997; Snaith and Haley, 1999). In our study,we proposed different facets to capture thistourism attitude, e.g. the emphasis on furtherencouraging tourism, the development ofadditional tourism facilities, the role oftourism in community, the valuation oftourists and contacts with them.

Hypotheses

Our basic research question raised in thisstudy is:

● whether a social distance scale can bedeveloped and purified to be used in thenew context of tourism research. As someevidence of reliability it is asked:

● if the end result of the development of asocial distance measure in two differentcountries will contain similar scale items.

Most social impact studies in tourism fol-low the paradigm of intrinsic and extrinsicinfluential dimensions, as proposed byFaulkner and Tideswell (1997). The majorityof investigations centre around communitycharacteristics and a variety of socio-demo-graphic indicators representing the individualresident. Some research papers try to addressmore explicitly the relationship between thelocal population and the tourism system, e.g.investigating the dependency on tourism(Smith and Krannich, 1998), communityinvolvement in tourism development(Bachleitner and Zins, 1999), or the residents’sentiment toward the community (Jurowski,1998). By considering the concept of socialdistance to tourists, we add a new perspectiveto the intrinsic dimension. Therefore, the fol-lowing model with an inter-attitudinal struc-ture is put forward when stating:

● that there is a positive relationshipbetween social distance and general atti-tudes toward tourism impacts.

In addition, it is maintained:

● that there is a positive relationshipbetween social distance and perceivedtourism impacts on the one side and resi-dent views regarding future tourism devel-opment on the other.

A Measurement Instrument for SocialDistance

Item collection (New Zealand)

To identify appropriate scale items to beincluded in a tourism social distance scale,focus groups and personal interviews wereundertaken in New Zealand, in regions cho-sen for their different intensities of tourismdue to their differing stages of tourism devel-opment. Additionally, regions were chosenbecause of the types of tourists they attracted;this was to get a range of destinations popularwith both domestic and international visitors.This stage of research was predominantlyundertaken to generate a list of contact situa-tions between the host and guest. Participantswere asked to list the types of contact situa-tions that can occur between the host andguest. To ensure an exhaustive list, the result-ing items were compared to situations dis-cussed in the literature. In total, 68 contactsituations were obtained (see Appendix TableA for a list of the 68 contact situations).

Judge ranking (New Zealand)

The methodology behind the design of thesocial distance scale was very similar to thatimplemented by Bogardus (1933). The 68contact situations were individually listed,typed, randomized by computer and copiedon to separate cards. These cards were givento 20 judges (people with a variety of back-grounds, all over 20 years of age, and a mix ofmen and women) who were asked to placethe statements into seven different pilesaccording to the intimacy of the specific situa-tion. The judges were told to imagine thecontact situation to be between a visitor to acommunity and the host, and requested torank the situation from 1 (most intimate) to 7(least intimate). Seven degrees of intimacywere chosen in line with the original methodused by Bogardus. The judges were also askedto ensure that each pile included at least fivestatements. This was to make certain thateach of the seven different degrees of inti-macy or social distance were accounted for.



The judges were subsequently asked to gothrough each pile and shift any statementsthat they felt would be better suited in a dif-ferent pile. They were asked to repeat thisprocedure until they were entirely satisfiedwith their choices.

Once all 68 items were sorted into sevenpiles, the pile number for each statement(1–7) was entered into the statistical com-puter package SPSS and the mean, mode, stan-dard deviation and range for each statementwas calculated (see Appendix Table A).

Although the design of the social distancescale in this research is very similar to that ofBogardus (1933), a significant departure isthe use of the Thurstone (1928) equal-appearing interval technique (adapted bySartain and Bell, 1949, for the use in socialdistance), i.e. the mean, mode and standarddeviation were all considered in the decisionof which items to include in the final scale, asopposed to the design by Bogardus, whichjust considered the mean. This was done toensure a more rigorous method of scaledevelopment.

The mean, mode and standard deviationof each situation was calculated to establishthe stage of the social distance scale eachitem would best fit (1–7). Items that had the

closest means and modes were chosen asgood, reliable representations of particularscale stages. For example, the item ‘Haveinto my home as dinner guests’ had a modeof 2, a mean of 1.85 and a low standard devi-ation (0.37); therefore, the judges were pre-dominantly placing it in the second stage ofthe social distance scale. An additional char-acteristic that was considered when choos-ing the items was their relationship totourism. For example, the item ‘Have aslong-term guests into my house, for exam-ple as exchange students’ (mean 1.15 andmode 1), seemed more relevant to tourismthan did ‘Marry or live with them’, whichactually had a mean of 1 and a 0.00 stan-dard deviation.

Table 4.2 shows the items that were mostfrequently placed in each pile (as shown bythe modes). In addition, the mean andmodes of these items were very closetogether.

Judge ranking (Austria)

In order to prepare for the scale purificationin Austria, the 68 statements from the NewZealand collection were semantically analysed.


Table 4.2. Items chosen as representations of stages in the traditional social distance scale (NewZealand). A yes/no response was requested.

Mean Mean Mode Mode SD SD

(initial 2nd (initial 2nd (initial 2nd No. Variable rank) rank rank) rank rank) rank

1 Would like having them into their house as long-term guests, e.g. as exchange students 1.15 1.05 1 1 0.37 0.22

2 Would like having them into their house as dinner guests 1.85 2.05 2 2 0.37 0.22

3 Would give a ride to them if they were hitchhiking/Later replaced by:Would enjoy having them temporarily work at

their place of employment 3.05 2.85 3 3 0.89 0.494 Would like sitting beside them on chair lift at a

ski field in New Zealand 4.25 4.05 4 4 0.97 0.225 Would enjoy sharing the bus into town with 5.20 5.25 5 5 0.83 0.446 Would like seeing them at our public swimming

pool 5.80 5.70 6 6 0.89 0.577 Would like seeing them in a souvenir shop 6.65 7.00 7 7 0.49 0.00

SD, standard deviation.


Some items were dropped due to inapplicabil-ity; others were slightly changed in the word-ing or for the context. ‘To be addressed in aforeign language’ was used instead of ‘Havingsigns in foreign language around our commu-nity’, ‘Using the water slide beside me’ insteadof ‘Share a jet boat ride with’. ‘Seeing themvisit the local church’ was added as one inter-action touching the residents’ socio-culturallife.

The task of sorting the remaining 59 cardsinto seven piles was kept identical to the proce-dure outlined in the previous sub-section. Asample of 32 judges of mixed age, balancedgender and different tourism experiences wasused to rank the written host–guest contact sit-uations. The results of this task are provided inAppendix Table A, together with the NewZealand rankings. Overall, there is high con-cordance among the judgements comparingthe means and the modes. Significant differ-ences can be defined as the cases where meansand/or modes deviate more than one scaleunit. This occurs ten times out of 56 commonitems. In general, these situations describecontacts in shops, restaurants or on leisureoccasions which are rated closer by Austrianthan by New Zealand judges. However, thispattern cannot be generalized, as similar situa-tions show much more congruence.

Seven statements were selected whichseemed to be good representatives for differ-ent levels of a closeness–distance continuum.Again, those items were taken with a low stan-

dard deviation, an appropriate mode valueand the lowest difference between mean andmode. In addition, the appropriateness forthe tourism context was considered. Theselected statements are listed in Table 4.3.

Judge re-ranking (New Zealand, Austria)

The consistency of the placements of thesechosen items was then checked; the sevenstatements were again randomized and givenback to the original 20/10 judges (NZ/A),who were asked to again rank them frommost intimate to least intimate. This was toensure consistency between the final sevenstatements and how they were ranked ini-tially. Tables 4.2 and 4.3 also show the resultsof this second ranking.

The ranking of the statements remainedconsistent. However, there were a number ofjudges unhappy with the inclusion of thethird item ‘Give a ride to if they were hitch-hiking’ (NZ). Many commented that theyfound this the most difficult item to rank,mainly because they never stopped to pick upany type of hitchhiker. Therefore it wasdecided that this item would be replaced with‘Having them work at my place of employ-ment’ which had a mean of 3.00 and a modeof 3 (in the original ranking). These state-ments were again given to the judges for theircomments and it was agreed unanimouslythat this item was more appropriate for the


Table 4.3. Items chosen as representations of stages in the traditional social distance scale (Austria). Ayes/no response was requested.

Mean Mean Mode Mode SD SD

(initial 2nd (initial 2nd (initial 2nd No. Variable rank) rank rank) rank rank) rank

1 Would like having them into their house as long-term guests, e.g. as exchange students 1.58 1.90 1 1 1.09 1.52

2 Having them work at my place of employment 2.74 2.60 2 2 1.32 1.583 Sit beside on the daily rail trip for work 3.71 3.80 3 3 1.49 1.404 Giving directions to them 3.90 3.90 4 6 1.54 1.915 Notice them watching us participate in our hobby,

for example playing cards, fishing 5.07 4.50 5 5 1.67 0.976 Seeing them at our public swimming pool 5.00 5.30 6 6 1.72 1.167 Have to walk around them standing in the

middle of a supermarket aisle 5.68 6.40 7 7 1.70 1.35

SD, standard deviation.


scale. In the Austrian sample, the rank orderof the seven statements remained the same.Some inconsistencies occurred with the item‘Giving directions to them’ resulting in amode value of 6. However, as no clear candi-date for replacing this particular statementexisted, it was kept in the measurementinstrument.

When comparing the re-ranked itemsbetween the two countries (Tables 4.2 and4.3), it can be observed that four out of sevenstatements are identical, having the same(items 1 and 6), almost the same (items 3/NZand 2/A), or a different (items 5/NZ and3/A) rank position. The last statements bothrefer to loose contact situations in a shop.The dinner guest situation (NZ) as well as the‘giving directions’ statement (A) have noequivalent in the other country scale.

Empirical Studies

Description for New Zealand

The resulting scale was empirically testedwithin a survey of New Zealand residents’ atti-tudes towards tourism. Forty-three items wereincluded to measure attitudes towards tourismboth in the respondent’s own community (forexample, ‘because of tourism, we have betterrecreational opportunities available in ourtown’) and at a national level (for example,‘tourism has improved the quality of service inshops, restaurants and hotels in New Zealand’).See Lawson et al. (1998) for a description of thedesign of these 43 items. These attitudinalstatements were later factor analysed to enablea correlation analysis between the social dis-tance dimensions and general attitudes towardstourism. The results of this test will be dis-cussed later in this chapter.

The questionnaires were distributedthroughout New Zealand in 19 destinationschosen to represent the stages of Butler’s(1980) Destination Life Cycle. For the socialdistance scale, respondents were given ayes/no response format, adapted from Lee etal. (1996) (Table 4.2). They were asked toindicate what they believed the average NewZealander’s reaction would be to findingthemselves in each of the different situations

with different tourist nationalities. This third-person, projective technique was adoptedbecause of the racial sensitivity of the ques-tions being asked, and it was thought thatrespondents may be more likely to projecttheir true feelings on to this ‘average’ NewZealander, rather than merely offering sociallyacceptable responses (Aaker et al., 1998).

Four different international tourist originswere chosen to test the proposition that theextent of social distance observed would varydepending on the cultural and physical simi-larity of the host and visitor. Germany, USA,Australia and Japan were deemed to repre-sent a range of cultures and physical charac-teristics, while also being of strategicimportance to New Zealand as markets forinbound tourists (www.tourisminfo.govt.nz).Different paired combinations of these fourcountries were included in the question-naires, resulting in two data sets on whichsubsequent analysis will be based: data set 1referred to Australian and German touristsand data set 2 referred to US and Japanesetourists. In total, 1193 surveys were distrib-uted, of which 518 usable questionnaires werereturned (an effective response rate of 43%).

Of the questionnaires returned, there wasa total of 1036 responses to the social distancescale (as two scales were included in each sur-vey). Because the social distance scale wasdesigned as a cumulative scale (Guttman,1944), the first part of the analysis involvedchecking the scale’s internal consistency.

Description for Austria

The Austrian validation of the social distancescale was done on a rather small sample, lim-ited to a region in the eastern part of Austriaclose to the Hungarian border. This area, cov-ering about 13 communities, is thinly popu-lated and at the very beginning of a dynamictourism development based on a new spa andgolf infrastructure. Oral interviews were con-ducted with 238 inhabitants representing theage, gender and professional structure of thestudy region. For measuring social distance,respondents were asked which contact situa-tions between themselves and tourists theywould accept without reluctance and which



they would rather like to avoid. A differentia-tion between nationalities was not considered,as the study area’s tourism experience is verymuch limited to Austrian and German touristsspeaking more or less the same language. Toexamine the various dimensions of tourismimpacts, respondents rated 21 statements on aseven-point Likert-scale (1 ‘fully applies’ to 7‘does not apply at all’). Additionally, 12 state-ments were formulated for capturing attitudestoward future tourism development.

Results

Social distance – internal consistency of NewZealand scale

To begin, the number of scale items that hadbeen ‘misranked’ was calculated, i.e. becausethe scale was cumulative, if a respondentanswered ‘yes’ to item ‘1’, they should havecontinued with ‘yes’ to all other items on thescale. Additionally, if the respondentanswered ‘no’ to question 1 and question 2,but ‘yes’ to question 3, he or she should con-tinue answering ‘yes’ on the remaining items.Thus, it was important to determine the num-ber of abnormal switches in the data (that is,answering ‘yes’ at the beginning of the scale,then shifting to ‘no’, or switching more thantwo times in the scale).

Such misranking was ascertained manuallyby printing out the raw data and checkingeach response for validity. From this, it wasobserved that item 3 on the scale (‘Wouldenjoy having them work at my place ofemployment’) caused a large number of mis-rankings; respondents may have indicated‘yes’ for items 1 and/or 2, but ‘no’ for item 3.Of the 1036 scales completed, the percentageof misrankings was 35% with item 3 includedbut only 16% when item 3 was excluded. Itappeared that respondents saw this item as areflection of New Zealanders missing out onjobs that ‘foreigners’ had taken. This idea wasexpressed in the comments volunteered atthe end of the questionnaire and had alsobeen alluded to in the exploratory focusgroups. For example, some respondentspointed out that Queenstown (one of NewZealand’s most popular international tourist

spots) has a lot of foreign workers, particu-larly of Japanese origin, in souvenir shops.When the judges used in the original scaledevelopment were approached and asked tocomment on this finding, they were not sur-prised at the way respondents to the question-naire appear to have interpreted this item.This may be one limitation of the traditionalscale’s design. That is, when judges rejectedthe third scale item originally proposed(‘Would give a ride to if they were hitchhik-ing’), they were only given one other option(‘Would enjoy having them work at my placeof employment’) which they obviously pre-ferred over the previous item, but there maystill have been a better item to choose.

Thus, the third item was excluded fromfurther analysis on the social distance scale,with it then becoming a six-point scale.Additionally, all other cases that changedtheir response more than once or began witha ‘yes’ and changed to ‘no’ were alsoexcluded from further analysis.

Another measure used to determine theinternal consistency of the social distancescale, after item 3 was excluded, was the criti-cal ratio (CR) of the remaining items(Guttman, 1950). Within the data set,responses on German and Japanese touristswere placed together as were responses toAustralian and US tourists. ForGermany–Japan together (507 validresponses), only 220 errors could be calcu-lated which results in a CR of 92.8 which isabove the critical threshold of 90. Taking thedata for Australia–US together (512 validresponses), only 162 errors occurred meaninga CR of 94.7. Taking all responses together(1014 valid responses), the errors were nomore than 382 yielding a CR of 93.7. Thisresult is more than satisfactory. If analysedindependently the CR for each nationality is95.2 for Australia, 94.1 for Germany(Spearman correlation of social distance: 0.65at P <0.01), 94.2 for the USA and 90.6 forJapan (Spearman correlation of social dis-tance: 0.54 at P <0.01). Whereas the measuresfor the distance to two foreign nationalitiesseem to be fairly correlated, the average rankorder is significantly different resulting in amuch higher distance to German or Japanesetourists (Mann–Whitney U test: P <0.05).



Social distance – internal consistency ofAustrian scale

Taking the initial order of items within thesocial distance scale resulted in a CR of 81.7.After closer investigation it turned out thatitem 5 (‘Notice them watching us participatein our hobby, for example playing cards’)produced the majority of errors, so it waseliminated from the scale. For furtherimprovement of the scale consistency, state-ment 4 (‘Giving directions to them’), whichwas not in a clear order position during there-rank task, had to be re-placed at the bot-tom of the scale, representing the least insult-ing situation. The new item hierarchygenerated a satisfying level for the CR of 94.8.

When comparing the social distance mea-sures between the two studies, similar distribu-tions can be observed. Less than 40% wouldaccept the closest contact situations totourists. Almost 50% (the German–Japaneserelationship only 37%) agree contacts up tothe second highest level. Altogether the per-ceived distance to Australian and US touristsdoes not deviate from that expressed byAustrian residents toward foreign tourists ingeneral (Mann–Whitney U test). Though, thedistribution of the measure shows a differentpattern considering the distance perceived forGerman–Japanese tourists giving a muchhigher distance (Mann–Whitney U: P <0.05).

Social distance – external validity

In a next step, the correlation between socialdistance and attitudes toward tourism is inves-tigated as outlined in the hypothesis section.For this purpose, the attitudinal statementswere first condensed by principal compo-

nents analyses. For the New Zealand sample,one dimension capturing the emphasis onlimiting growth could be identified from arather small item pool (Table 4.4). Themajority of statements covered the variousimpacts perceived by the residents. From thisscope of attitudes, five dimensions wereextracted (Table 4.5), which are labelled as:improvements in lifestyle, negative environ-mental impacts, negative socio-culturalchange, negative price effects and avoid con-tacts to tourists. The factors extracted notmore than 52% of the variance. They servean exploratory purpose only which isreflected in the varying reliability ratio ofCronbach’s alpha.

From the Austrian data set, two dimen-sions were extracted representing attitudestoward future tourism development: activetourism encouragement and tourism in thecommunity (Table 4.6). The tourism impactattitudes were reduced by principal compo-nents analyses. Five dimensions count for67% of the variance in the statements listedin Table 4.7: negative social impacts, eco-nomic benefits, improvements for the com-munity, individual benefits and negativeecological impacts.

The final investigation focused on the rela-tionship between the social distance measuresand the attitudes toward tourism. For this pur-pose, correlation analyses were conducted(employing Spearman-type for the social dis-tance and Pearson-type for inter-attitudinalrelationships). For the New Zealand samplesignificant correlations between the two socialdistance measures and the perceived tourismimpact factors could not be detected. Theonly significant link appeared between theGermany–Japan related social distance to the‘Limiting growth factor’ (correlation = 0.11; P


Table 4.4. Attitude towards future development of tourism (New Zealand).

LimitedStatements and factors: attitudes towards tourism development growth

I prefer to have as little contact as possible with overseas 0.79It would be better to have fewer tourists and less money from tourism if it meant that

we could have free access to New Zealand 0.77I would like to see an increase in the number of tourists in our community �0.73Cronbach’s alpha 0.62



Tab

le 4

.5.

Atti

tude

tow

ards

tour

ism

impa

cts

(New

Zea

land

).

Sta

tem

ents

and

fact

ors:

atti

tude

s to

war

ds

Impr

ovem

ents

Neg

ativ

eA

void

Neg

ativ

eN

egat

ive

tour

ism

dev

elop

men

tin

life

styl

een

viro

nmen

tal

cont

act

pric

e ef

fect

sch

ange

The

cur

rent

leve

l of t

ouris

m h

as im

prov

ed e

mpl

oym

ent o

ppor

tuni

ties

in th

is to

wn

0.77

Tour

ism

enc

oura

ges

a w

ide

varie

ty o

f cul

tura

l act

iviti

es (

e.g.

cra

fts, a

rts,

mus

ic e

tc.)

in th

is to

wn

0.72

Sho

ppin

g in

this

tow

n (e

.g. c

hoic

e of

sho

ps a

nd lo

nger

ope

ning

hou

rs)

is b

ette

r be

caus

e of

tour

ism

0.71

Mos

t peo

ple

in th

is to

wn

are

bette

r of

f bec

ause

of t

ouris

m0.

68To

uris

m h

as m

ade

loca

ls m

ore

prou

d of

thei

r co

mm

unity

0.67

Thi

s ar

ea h

as m

ore

nigh

tlife

(e.

g. n

ight

club

s an

d re

stau

rant

s) b

ecau

se o

f tou

rism

0.62

The

qua

lity

of p

ublic

ser

vice

s lik

e w

ater

, sew

erag

e an

d pu

blic

tran

spor

t in

this

tow

n ha

s im

prov

ed d

ue to

tour

ism

0.49

The

ben

efits

from

tour

ism

are

dis

trib

uted

fairl

y th

roug

hout

our

com

mun

ity0.

49T

he tr

affic

aro

und

tow

n of

ten

beco

mes

con

gest

ed b

ecau

se o

f tou

rist t

raf fi

c0.

79To

uris

m h

as c

reat

ed p

arki

ng p

robl

ems

in o

ur c

omm

unity

0.75

Tour

ism

has

incr

ease

d cr

ime

in o

ur c

omm

unity

0.

67T

here

is m

ore

litte

r in

my

com

mun

ity b

ecau

se o

f tou

rism

0.54

I avo

id p

lace

s w

here

I kn

ow th

ere

will

be

a lo

t of t

ouris

ts0.

75I g

et fr

ustr

ated

with

the

way

tour

ists

driv

e0.

62O

vers

eas

tour

ists

sho

uld

pay

mor

e th

an lo

cal r

esid

ents

to v

isit

Nat

iona

l Par

ks a

nd o

utdo

or

recr

eatio

n fa

cilit

ies

0.55

Tour

ism

mak

es p

rices

ris

e so

New

Zea

land

ers

can

no lo

nger

af fo

rd to

hol

iday

in c

erta

in a

reas

of

New

Zea

land

0.

71Lo

cal p

eopl

e ne

ed to

hav

e m

ore

inpu

t int

o to

uris

m d

evel

opm

ent

0.47

New

bui

ldin

gs, f

rom

tour

ism

, hav

e ch

ange

d th

e ap

pear

ance

of o

ur to

wn

for

the

wor

se0.

79O

ur w

ay o

f life

in th

is to

wn

has

chan

ged

to s

uit t

ouris

ts0.

73B

ecau

se o

f tou

rism

, I s

omet

imes

feel

like

a s

tran

ger

in m

y ow

n to

wn

0.50

Cro

nb

ach

’s a

lph

a0.

800.

720.

460.

480.

62


< 0.05), i.e. the greater the social distance per-ceived toward tourists from Germany or Japanthe more adverse is the opinion toward futuretourism development. The inter-attitudinalcorrelations have a range from 0.18 to 0.30and show the expected direction: the moreindividual benefits perceived the less reluctantthe residents are toward future growth intourism and the more adverse effects they per-ceive (price rise, environmental damages,social change) the less open they are forfuture tourism development.

For the Austrian data set, the sameapproach was applied. The significant correla-tions are summarized in Fig. 4.1 which basicallyshows that the higher the social distancetoward tourists is perceived the more negativethe impacts on the environment, the lessimprovements for the community and the lessindividual benefits are attributed to the tourismactivity. Hence, the social distance measureinfluences the attitude toward future tourismdevelopment directly and indirectly throughsome aspects of perceived effects spreadingfrom the tourism activities. Interestingly, no sig-nificant correlation could be observed betweensocial distance and perceived negative socialimpacts. The negative impacts (social and eco-logical) are more related to the tourism devel-opment of the community, whereas the positiveand more economic aspects are rather con-nected to the mind-set toward the generaltourism encouragement.

Discussion and Implications

This chapter describes the first measure-ment of host attitudes towards differenttourist nationalities and attempts to link thefindings to general attitudes towards tourismimpacts and development. The concept ofsocial distance is used to measure this con-struct and two scales are developed, usingthe same methodology (adopted fromBogardus, 1933).

One of the most important dimensions ofthe tourism product is the host communityand their interaction with the tourist. Thesocial distance concept suggests that the hostcommunity may have quite different degreesof tolerance towards different tourist nation-alities, and it is suggested that these differ-ences may be reflected in the hostcommunity’s behaviour towards tourists. Asocial distance scale can be used to measurethese tolerance degrees, which in turn canaid tourism developers and planners in suchdecisions as target market strategies.

The development of the social distancescales in both New Zealand and Austria showsa number of commonalities; for example themean and mode of many of the itemsbetween the two countries are very similar.Although the final scores are not identical,this appears to have more to do with differ-ences between the samples used in the study,rather than an inadequacy with the scale


Table 4.6. Attitudinal factors towards future tourism development and loading items (Austria).

Active tourism Tourism in the Statements and factors: attitude towards tourism development encouragement community

Tourism should be encouraged much more in this region 0.82Tourism should be encouraged actively in this community 0.79The community should create additional tourism facilities 0.75I favoured the development of the spa and the golf course from

the very beginning 0.70The communities in this region should not try to attract more tourists �0.70I prefer not to familiarize with tourists �0.53Tourism entails that leisure facilities are crowded �0.51Our community should be transformed to a tourism destination 0.73Tourism should play a major role in this community 0.67To have a job in tourism is very much appreciated 0.67Tourists are valuable 0.63Cronbach’s alpha 0.77 0.74



Tab

le 4

.7.

Atti

tudi

nal f

acto

rs to

war

ds to

uris

m im

pact

s an

d lo

adin

g ite

ms

(Aus

tria

).

F1

F2

F3

F4

F5

Neg

ativ

eN

egat

ive

Eco

nom

icIm

prov

emen

ts –

ec

olog

ical

Indi

vidu

alA

ttitu

des

tow

ards

tour

ism

impa

cts

soci

al im

pact

sbe

nefit

sco

mm

unity

impa

cts

bene

fits

Loca

ls a

re d

iscr

imin

ated

aga

inst

in r

esta

uran

ts c

ompa

red

to to

uris

ts0.

86To

uris

ts a

re m

ore

valu

ed in

som

e sh

ops

than

loca

ls0.

82To

uris

ts c

onfli

ct w

ith c

omm

unity

life

0.67

Con

flict

s be

twee

n to

uris

ts a

nd lo

cals

ofte

n ar

ise

0.64

Tour

ism

has

incr

ease

d cr

ime

in m

y co

mm

unity

0.49

Due

to th

ese

faci

litie

s m

ore

tour

ists

are

attr

acte

d to

this

reg

ion

0.79

The

spa

and

gol

f cou

rse

bene

fit th

e w

hole

reg

ion

0.79

The

ben

efits

of t

ouris

m o

utw

eigh

the

nega

tive

cons

eque

nces

of t

ouris

m d

evel

opm

ent

0.62

Tour

ism

dev

elop

men

t in

my

com

mun

ity w

ill p

rovi

de m

ore

jobs

for

loca

l peo

ple

0.60

The

qua

lity

of p

ublic

ser

vice

s ha

s im

prov

ed d

ue to

tour

ism

in m

y co

mm

unity

0.87

Sho

ppin

g op

port

uniti

es a

re b

ette

r in

my

com

mun

ity a

s a

resu

lt of

tour

ism

0.86

Com

mun

ity r

oads

hav

e im

prov

ed d

ue to

the

tour

ism

dev

elop

men

t0.

59A

gric

ultu

ral a

ctiv

ities

are

red

uced

due

to th

e sh

rinki

ng a

vaila

ble

spac

e0.

59To

uris

m h

as n

egat

ivel

y im

pact

ed th

e en

viro

nmen

t0.

77T

here

is m

ore

litte

r in

my

com

mun

ity fr

om to

uris

m0.

81I h

ave

mor

e m

oney

to s

pend

as

a re

sult

of to

uris

m0.

87To

uris

m in

my

com

mun

ity h

as in

crea

sed

my

stan

dard

of l

ivin

g0.

86C

ron

bac

h’s

alp

ha

0.78

0.71

0.76

0.76

0.65


itself. The internal consistency according toGuttman’s proposal is very satisfactory aftereliminating one item of the initially con-ceived seven-point scale. While with the NewZealand data the CR could not be improvedby re-ordering the remaining items, one re-placement within the Austrian scale increasedthe CR substantially. This weakness may bedue to unforeseen differences in the percep-tions and tourism experiences of the judges(for the scale development) and the residents(in the validation study).

Despite the fact that some semantic differ-ences are inherent between the social distancescales developed in New Zealand and Austria,the distributions of the sample populationsacross the scale continuum support the follow-ing consideration. There are obvious differ-ences referring to the stage in the tourismdestination lifecycle. The New Zealand socialdistance scale was designed and tested by sam-

ples throughout New Zealand; regions repre-senting various degrees of tourism develop-ment. This will have an impact on attitudes,particularly since some regions are more pop-ular with the obviously culturally differenttourists. The Austrian scale on the other handwas developed and tested on one region ofresidents which is in an early stage of tourismdevelopment. The contact situations with for-eign cultures are quite distinct. Due to the his-toric settlement and border evolution, theexchange with Hungarian and Croatian peo-ple as permanent residents is for a good partof the locals more intense than the experi-ences with Austrian, German or guests fromother countries. Nevertheless, no significantdifferences occurred when comparing thesocial distance perceived by New Zealanderstoward Australian and American tourists, andthat perceived by (eastern) Austrians towardany tourists. Yet, the social distance scale


–.39

Social distance

Neg. socialimpacts

Negativeecologicalimpacts

Tourism in thecommunity–.34

Improvementsfor

community

Active tourismencouragement

.38

.49

.32

.40

Attitude towardstourism development

Economicbenefits

Individualbenefits

Tourism impacts andattitude scale

–.13

.15

–.14

–.19

Arrows = Correlations

Social distance(Guttman scale)

Fig. 4.1. Correlations between social distance and tourism attitudes (Austria).


discriminated very well between the English-speaking nationalities and the culturally moredifferent German and Japanese tourists, indi-cating a significant lower versus higher dis-tance.

Put together, it can be derived that thesemantic differentiation of the scale itemsdepends on the background and tourismrelatedness of the target population. The dis-criminating power of the scale, as well as thecommon mean for the New Zealand andAustrian sample population, can be taken asevidence that the construct is distinct andvalid. However, it does not seem to be com-pletely independent from the stage oftourism development as the correlations withthe attitudes toward tourism impacts show avery different pattern between the two coun-tries. Another difference in the scale applica-tion stems from the perspective of distanceattribution. While in the New Zealand survey,respondents were asked a projective questionof social distance, the Austrian survey directlyreferred to the respondent’s perception ofdistance. However, how and why these differ-ences occur should be investigated more sys-

tematically, i.e. with more control variableswith regard to regions and cultural differ-ences between hosts and guests.

Future research recommendations includethe construction of a multi-item social distancescale using for example Likert-type scales foreach item/statement. Most contemporaryresearchers now favour the use of the multi-item measures, which allow for rigorous reliabil-ity testing and the examination of some aspectsof validity using empirical methods (Churchill,1979). A multi-item scale may have been moreapplicable to the attitude statements includedin the surveys and provided more statisticallysignificant results. On the other hand, a reliablelean Guttman-type scale means more efficiencyand more time left for other survey topics whenconducting research studies.

Acknowledgements

The authors thank Professor Rob Lawson andDr Sarah Todd for their assistance in this pro-ject and the Public Good Science Fund forfunding the research in New Zealand.


References

Aaker, D.A., Kumar, V. and Day, G.S. (1998) Marketing Research. John Wiley & Sons, New York.Anderson, M.J. (1983) Prejudice in Dunedin: an examination of prejudice in Dunedin using the social dis-

tance technique. Unpublished MArts, University of Otago, Dunedin, New Zealand.Ap, J. and Crompton, J. (1998) Developing and testing a tourism impact scale. Journal of Travel Research 37,

120–130.Bachleitner, R. and Zins, A. (1999) Cultural tourism in rural communities: the residents’ perspective.

Journal of Business Research 44, 199–209.Bogardus, E.S. (1922) A History of Social Thought. University of Southern California Press, Los Angeles,

California.Bogardus, E.S. (1925) Social distance and its origins. Sociology and Social Research 9, 216–225.Bogardus, E.S. (1933) A social distance scale. Journal of Applied Psychology 17, 265–271.Bogardus, E.S. (1934) Social distance in Shakespeare. Sociology and Social Research 18, 67–73.Bogardus, E.S. (1940) Scales in social research. Sociology and Social Research 24, 69–75.Bogardus, E.S. (1967) Measuring Social Distances. Readings in Attitude Theory and Measurement. John Wiley &

Sons, New York.Butler, R.W. (1980) The concept of a tourist area cycle of evolution: implications for management of

resources. Canadian Geographer 24, 5–12.Churchill, G.A., Jr (1979) A paradigm for developing better measures of marketing constructs. Journal of

Marketing Research 16, 64–73.Crull, S.R. and Bruton, B.T. (1979) Bogardus’ social distance in the 1970s. Sociology and Social Research 63,

771–783.Dyer, J.A., Vedlitz, A. and Worchel, S. (1989) Social distance among racial and ethnic groups in Texas:

some demographic correlates. Social Science Quarterly 70, 607–616.Faulkner, B. and Tideswell, C. (1997) A framework for monitoring community impacts of tourism. Journal

of Sustainable Tourism 5, 3–28.


Getz, D. (1994) Resident’s attitudes towards tourism. Tourism Management 15, 247–258.Guttman, L. (1944) A basis for scaling qualitative data. American Sociological Review 9, 139–150.Guttman, L. (1950) The basis for scalogram analysis. In: Stouffer, S.A., Guttman, L., Suchman, E.A.,

Lazarsfeld, P.F., Star, S.A. and Claussen, J.A. (eds) Measurement and Prediction, Vol. 4. PrincetonUniversity Press, Princeton, New Jersey, pp. 60–90.

Hagendoorn, L. and Kleinpenning, G. (1991) The contribution of domain-specific stereotypes to ethnicsocial distance. British Journal of Social Psychology 30, 63–78.

Hraba, J., Brinkman, R. and Grayray, P. (1996) A comparison of black and white prejudice. SociologicalSpectrum 16, 129–157.

Hraba, J., Hagendoorn, L. and Hagendoorn, R. (1989) The ethnic hierarchy in The Netherlands: socialdistance and social representation. British Journal of Social Psychology 28, 57–69.

Jacoby, J. (1978) Consumer research: a state of the art review. Journal of Marketing 42, 87–96.Jaspars, J. and Hewstone, M. (1982) Cross-cultural interaction, social attribution and inter-group relations.

In: Bochner, S. (ed.) Cultures in Contact: Studies in Cross-Cultural Interaction. Pergamon Press, Oxford,pp. 127–156.

Jurowski, C. (1998) A study of community sentiments in relation to attitudes toward tourism development.Tourism Analysis 3, 17–24.

Lankford, S.V. and Howard, D.R. (1994) Developing a tourism impact attitude scale. Annals of TourismResearch 21, 121–139.

Lawson, R.W., Williams, J., Young, T. and Cossens, J. (1998) A comparison of residents’ attitudes towardstourism in 10 New Zealand destinations. Tourism Management 19, 247–256.

Lee, M.Y., Sapp, S.G. and Ray, M.C. (1996) The reverse social distance scale. The Journal of Social Psychology136, 17–24.

Lemaine, G. and Ben Brik, J. (1997) Rejection. From attitudes to intentions to exclude: social distance,phenotype, race and culture. Social Science Information 1, 81–113.

Lindberg, K. and Johnson, R. (1997) Modeling resident attitudes toward tourism. Annals of Tourism Research24, 402–424.

Nix, J.V. (1993) Assessing the Existence of Social Distance and Factors that Affect its Magnitude at a SouthernUniversity. Department of Sociology, University of Mississippi.

Owen, C.A., Eisner, H.C. and McFaul, T.R. (1981) A half-century of social distance research. Sociology andSocial Research 66, 80–98.

Park, R.E. (1924) The concept of social distance – as applied to the study of racial attitudes and racial rela-tions. Journal of Applied Sociology 8, 339–344.

Perdue, R.R., Long, P.T. and Allen, L. (1987) Rural resident tourism perceptions and attitudes. Annals ofTourism Research 14, 420–429.

Prothro, E.T. and Miles, O.K. (1953) Social distance in the Deep South as measured by a revised Bogardusscale. The Journal of Social Psychology 37, 171–174.

Rollins, R. (1997) Validation of the TIAS as a tourism tool. Annals of Tourism Research 24, 740–742.Sartain, A.Q. and Bell, H.V., Jr (1949) An evaluation of the Bogardus scale of social distance by the method

of equal-appearing intervals. The Journal of Social Psychology 29, 85–91.Schneider, I.E., Lankford, S. and Oguchi, T. (1997) The cross-cultural equivalence of the TIAS: summary

results. Annals of Tourism Research 24, 994–998.Smith, M.D. and Krannich, R.S. (1998) Tourism dependence and resident attitudes. Annals of Tourism

Research 25, 783–802.Snaith, T. and Haley, A. (1999) Residents’ opinions of tourism development in the historic city of York,

England. Tourism Management 20, 595–603.Thurstone, L.L. (1928) Experimental study of nationality preferences. The Journal of General Psychology 1,

405–425.Triandis, H.C. and Triandis, L.M. (1960) Race, social class, religion, and nationality as determinants of

social distance. Journal of Abnormal and Social Psychology 61, 110–118.Trlin, A.D. (1970) Social distance and assimilation orientation: a survey of attitudes towards immigrants in

New Zealand. Pacific Viewpoint 12, 141–162.Vaughan, G.M. (1962) The social distance attitudes of New Zealand students towards Maoris (sic) and fif-

teen other national groups. The Journal of Social Psychology 57, 85–92.Williams, J. and Lawson, R. (2001) Community issues and resident opinions of tourism in New Zealand.

Annals of Tourism Research 28, 269–290.www.tourisminfo.govt.nz (20 February 2003).Yoon, Y., Chen, J.S. and Gürsoy, D. (1999) An investigation of the relationship between tourism impacts

and host communities’ characteristics. Anatolia: an International Journal of Tourism and HospitalityResearch 10, 29–44.



Appendix

Table A. Mean, mode and standard deviation calculations for the 68 items.

New Zealand Austria

Social distance item Mean Mode SD Mean Mode SD

Marry or live with them 1.00 1 0.00 1.58 1 1.09Sexual relations with them 1.05 1 0.22 n.a.Having them as flatmates 1.45 1 0.51 2.10 1 1.96Have them into my house as long-term guests, 1.15 1 0.37 1.58 1 1.09

e.g. as exchange studentsHave them as friends or relatives 1.45 1 0.60 1.74 1 1.46Have them into my house as dinner guests 1.85 2 0.37 1.90 1 1.08Have as friends of my family into my house 1.75 2 0.64 1.74 1 1.29Have as neighbours 2.40 2 0.60 2.45 1 1.34Helping them in car accident 2.45 2 1.32 2.87 1 1.73Drink with them at the pub 2.65 2 0.75 3.00 2 1.61Have them share the same room as myself in 2.60 2 1.31 3.00 2 1.61

a hostelHave in to view my house or garden 2.80 2 1.06 2.68 1 1.90Seeing them have a car accident 4.15 2a 1.60 5.26 7 1.86Having their children come to the same 3.45 3 1.28 3.06 2 1.44

crèche my children useHaving them work at my place of employment 3.00 3 0.92 2.74 2 1.32Give a ride to if they were hitchhiking 3.05 3 0.89 3.13 2 1.67Meeting them at conferences 3.55 3 0.89 4.10 4 1.99Have them to stay at accommodation I own, 3.40 3 1.19 3.10 2 1.62

for example bed and breakfast or motelDance with them at a nightclub 2.70 3 1.08 3.32 3 1.89Sit beside on a 24-hour flight to London b 3.50 3 1.10 3.19 4 1.62Having them into our schools or other 4.30 3a 1.22 3.26 4 1.32

education institutionsSit beside on the same domestic flight within 4.00 4 1.03 n.a.

New ZealandGiving directions to them 4.75 4 0.91 3.90 4 1.54Sit beside on the Trans Alpine Train 4.05 4 1.05 4.87 5 1.48

(a scenic train) b

Sitting beside them on the bus into town 4.65 4 1.04 4.10 3 1.64Share a jet boat ride with, for example on the 4.35 4 1.18 n.a.

Shotover JetShare a chair lift with at one of our ski fields 4.25 4 0.97 n.a.Share a helicopter flight to Milford with 4.15 4 1.04 n.a.Meeting them on our national walking/ 4.40 4 1.10 5.23 4 1.54

tramping tracksHave them at the same accommodation that 5.00 4a 0.86 2.39 2 1.38

I stay atHaving my photo taken by them 4.60 4a 1.23 4.13 5 1.73Have them sitting behind me in a restaurant 5.70 5 0.86 5.06 5 1.57Share the bus into town with 5.20 5 0.83 3.71 3 1.49Notice them watching us participate in our 5.05 5 1.05 5.06 5 1.67

hobby, for example line dancingTaking photos of them 4.80 5 1.06 4.32 4 1.80

Continued



Table A. Continued.

New Zealand Austria

Social distance item Mean Mode SD Mean Mode SD

Having them share the same recreation 4.55 5 1.19 3.97 3 1.83activities as us, for example horse riding or boating on the Lake

Having them serve me in a shop in New Zealand b 5.15 5a 0.99 4.00 3 1.75Have to walk around them standing in the 6.05 6 0.89 5.68 7 1.70

middle of a supermarket aisleSee them at our local pub 5.35 6 1.14 4.39 4 1.52See them in the shops I frequently visit 5.80 6 0.83 4.55 5 1.46See them in the butchers I frequently visit 5.70 6 0.92 4.84 4 1.66Having signs in foreign languages around our 5.50 6 1.54 n.a.

communitySee them use our local park 5.90 6 0.91 n.a.Seeing them at our public swimming pool 5.80 6 0.89 5.00 6 1.71See them in the bakery I frequently visit 5.80 6 0.83 4.97 4 1.33See them dine at my local café 5.45 6 1.05 4.65 6 1.82Standing in line waiting behind them, for 6.00 6 0.97 5.48 7 1.79

example to get movie ticketsFollowing them driving a campervan on our roads 6.45 7 1.10 6.10 7 1.54Following them driving a rental car on our roads 6.55 7 0.69 n.a.Seeing them stopped in the middle of the road 6.65 7 0.67 5.58 7 1.82

to take photosSeeing them in big groups walking around town 6.15 7 0.99 5.77 7 1.38Having them cross the road anywhere 6.70 7 0.66 n.a.Seeing them walk on our beaches 6.25 7 1.12 n.a.Having them take photos of our farm animals 6.40 7 0.82 5.03 7 1.76Seeing them visit our tourist destinations when 6.00 7 1.17 4.45 3 1.73

I am there, for example Mount Cook or Milford Sound b

Seeing them in our local museums 6.25 7 0.79 5.03 7 2.09Seeing them fish in our waters 6.30 7 1.03 5.19 6 1.58Seeing them at tourist spots or attractions 6.45 7 0.83 n.a.

when I am also there, for example at Larnach’s Castle

See them at a nightclub 6.15 7 1.04 5.13 4 1.38See them in a souvenir shop 6.65 7 0.49 5.35 6 1.85Seeing them playing sport 6.55 7 0.60 5.61 7 1.61See them on the Esplanade/water front b 6.60 7 0.60 5.35 7 1.68Seeing them watching live sport 6.45 7 0.69 4.84 4 1.73See them at the service station 6.55 7 0.76 5.77 7 1.80See them at the bank 6.55 7 0.76 5.19 7 1.83Seeing them camp anywhere 6.65 7 0.67 5.48 7 1.79See tourists outside the Beehive 6.85 7 0.37 n.a.Seeing them at other tourist spots around the 6.75 7 0.79 5.77 7 1.76

world, for example tourists at Surfers Paradise, Australia b

To be addressed in a foreign language c n.a. 4.06 5 1.97Seeing them visit the local church n.a. 5.03 7 2.06Using the water slide beside me n.a. 4.73 7 2.02

a Multiple modes exist. The smallest value is shown. b Item slightly changed in Austrian study.c Item used in Austrian study only. n.a., not applicable or not used.



Chapter fiveMotivation for Domestic Tourism: a Case Study of theKingdom of Saudi Arabia

Naima B. Bogari,1 Geoff Crowther2 and Norman Marr2

1King Abdul Aziz University, Faculty of Economics and Administration, Women’s Campus, POBox 42804, Jeddah 21551, Saudi Arabia; 2Department of Marketing, University ofHuddersfield, Queensgate, Huddersfield HD1 3DH, UK

Abstract

Tourism motivation in developing countries and Islamic culture has received scant attention fromresearchers. The key to understanding tourism motivation is to see holiday travel as a satisfier of needs andwants. The literature on tourism often conceptualizes tourist motives in terms of push and pull forces. Theidea behind this concept is that people travel because they are pushed by their own internal forces andpulled by external forces of the destination attributes. One way to analyse travel motivation is to examinethe notion of push and pull demand stimulation. The objectives of this research were to understand bothpush and pull motivation for domestic tourism and the relationship between these motivations for Sauditourists. The findings indicated nine push factors: ‘cultural value’, ‘utilitarian’, ‘knowledge’, ‘social’, ‘eco-nomical’, ‘family togetherness’, ‘interest’, ‘relaxation’ and ‘convenience of facilities’, and nine pull factors:‘safety’, ‘activity’, ‘beach sports/activities’, ‘nature/outdoor’, ‘historical/cultural’, ‘religious’, ‘budget’,‘leisure’ and ‘upscale’. This study found that the most important push and pull factors as perceived bySaudi tourists are ‘cultural value’ and ‘religious’. The study also confirms the relationship between pushand pull factors.

Introduction

In Saudi Arabia, there is a growing amount ofleisure free time and a high percentage of dis-posable income is being spent on variousforms of tourism; such trends have increasedthe number of Saudis travelling to tourist des-tinations, internationally or domestically.Consequently, spending the annual holidayaway from Saudi is normal for most Saudi

families. It is estimated (Sajini, 1997) that thetotal expenditure on domestic tourism inSaudi Arabia is only 16.7% of total tourismexpenditure. International expenditure wasestimated (in 1995) to be million US$7.6 andincreased to more than million US$8.2 in1997 (Asharq Alawsat Newspapers, 1998), whichis about 17.3% from oil revenue estimated in1998 and nearly 5.6% of the Saudi G.D.P.According to Economist Intelligence Unit



Limited (1998), Saudis themselves spend ontourism ($17bn pa) and only a small percent-age is spent in the country. Foreign tourism isa substantial drain on the current account, sothe government has been trying to persuademore of its citizens to holiday at home.

Many Saudis take their annual holidaysabroad where better recreational opportuni-ties and facilities are identified to exist.Within Saudi Arabia, few facilities are offeredin the countryside and desert for campingand other activities. National parks are beingset up, but often with few attractions andfacilities for tourists. In the last 25 years thenumber of hotel rooms has doubled; most ofthese are suitable, or in an appropriate loca-tion, for holiday use (Al-Thagafy, 1991) butsatisfaction levels tend to be lower amongthose better-educated and wealthier respon-dents. Contributing to this is the lack of coor-dination between tourists needs (demand)and supply of facilities, which results from thedivided responsibilities of the governmentand private sectors providing tourism facili-ties (Al-Thagafy, 1991). Also, there is a lack ofresearch data about tourists’ needs.

Study Objectives

This chapter has two objectives. First, studieson tourist motivation in Saudi Arabia and theArab Muslim culture are very limited, thusthis study makes a contribution to cover thisgap. It assesses the motivation ‘push’ and‘pull’ factors of tourist behaviour towardsdomestic tourism in an Islamic culture. Thisis important for the tourism industry fromthe planning and marketing managementperspective in order to segment the marketeffectively. Second, the study examines thenature of the relationship between push andpull factors in domestic Saudi tourism.

Travel Push and Pull Motivation

The field of tourism motivation research con-tinues to attract increasing attention fromresearchers (Holden, 1999), though it hasgained importance particularly over the lastdecade. Murphy (1985) argues that motivation

is essential for the development of tourism,because, without interest or the need to travel,the tourism industry could not exist. Gilbert(1991) argued that an understanding of moti-vation is important since it forms the maininfluence of tourism demand patterns. In addi-tion, it is one of the most complex areas oftourism research (Sharpley, 1999). One of theearlier reasons for emphasizing the significanceof tourist motivation came from marketers andpromoters of tourism (Howells, 2000). Therehave been several motivational studies oftourism, including Plog (1974), Crompton(1979), Hudman (1980), Dann (1981), Iso-Ahola (1982), Beard and Ragheb (1983),Pearce and Caltabiano (1983), Mansfeld(1992), Uysal and Hagan (1993) and Fodness(1994). Such studies of tourism motivation arederived from a variety of disciplinary subjectsthat have led to a variety of approaches.

The key to understanding tourism motiva-tion is to approach vacation travel as a satisfierof needs and wants, and literature on tourismoften conceptualizes tourist motives in termsof push and pull forces. The idea behind thisconcept is that people travel because they arepushed by their own internal forces andpulled by external forces of the destinationattributes (Dann, 1977, 1981; Crompton,1979; Hudman, 1980; Oh et al., 1995; Balogluand Uysal, 1996; Kim and Lee, 2002). Dann(1977) identified a two-level scheme of factorsthat motivate travellers to travel and to go tospecific destinations; this two-level scheme inthe travel decision-making process isdescribed as ‘push’ and ‘pull’. Push factors aresocio-psychological motivations that predis-pose, while the pull factors are those thatattract a person to a specific destination whenthe decision to travel has been made (Oh etal., 1995). Dann (1977) suggested thatanomie and ego-enhancement were the travelmotives. Crompton (1979) agreed withDann’s fundamental idea of push and pullmotives. Crompton identified seven push andtwo pull motives for travel. The push motiveswere the escape from perceived mundaneenvironment, exploration and evaluation ofself, relaxation, prestige, regression, enhance-ment of kinship relationships and facilitationof social interaction, while the pull motiveswere novelty and education. Hudman (1980)

52 N.B. Bogari et al.


used Maslow’s hierarchy of needs as a basis forcategorizing push factors of travel.

Several empirical examinations of bothpush and pull factors have been carried out inthe travel and tourism literature (Uysal andJurowski, 1994; Oh et al., 1995; Baloglu andUysal, 1996; Hanqin and Lam, 1999; You et al.,2000; Kim and Lee, 2002). Some studiesexamined the push factors only (Figler et al.,1992; Fodness, 1994; Cha et al., 1995) or pullfactors only (Sirakaya and McLelland, 1997;Klenosky, 2002). These studies differ in theapproach employed to identify push and pullfactors; some have employed quantitativemethods and multivariate modes of analysis(Uysal and Jurowski, 1994; Oh et al., 1995;Baloglu and Uysal, 1996; Hanqin and Lam,1999; You et al., 2000; Kim and Lee, 2002),others are based on a qualitative approachsuch as personal interview (Crompton, 1979;Pearce and Caltabiano, 1983; Klenosky, 2002).Uysal and Jurowski (1994) examined thenature and extents of the relationshipbetween push and pull factors for pleasuretourism. They reported high associationsbetween push and pull factors in a canonicalcorrelation analysis. Oh et al. (1995) andBaloglu and Uysal (1996) employed a correla-tion analysis and identified a significant rela-tionship between push and pull factors. Theysuggested that examining push and pull moti-vation simultaneously would be useful in seg-menting markets, in designing promotionalprogrammes and packages and in destinationdevelopment decision making. Hanqin andLam (1999) adopted a model based on pushand pull factors as a conceptual framework intheir study, and the results indicate that theimportance of push and pull factors in moti-vating Chinese travellers is different from thatfound in other studies. The study found thatmainland Chinese tourists perceive ‘knowl-edge’ and ‘high-tech image’ as the mostimportant push and pull factors. You et al.(2000) used Dann’s push and pull theory as aconceptual framework to test if travellers fromthe UK and Japan had different travel motivesand benefit-seeking patterns. The major find-ing was that UK and Japanese long-haul trav-ellers differ significantly on both push andpull forces. The findings of Kim and Lee(2002) confirmed the results of the study by

Uysal and Jurowski (1994), who reported arelationship between push and pull factors byusing multiple regression analysis. From amarketing perspective, tourism products canbe designed and marketed as solutions to con-sumer’s needs (Fodness, 1994). One way todetermine travel motivation is to examine thenotion of push and pull demand stimulation.

Methodology

A survey using a self-administered question-naire was conducted to collect primary datafrom a convenient sample of those visiting thetwo cities in Saudi Arabia, Jeddah and Abha.The application of a convenience sample wasnecessitated by the problems of samplingbeing prevalent in the Saudi environment.Such problems have been extensively docu-mented in the literature (Yavas et al., 1987;Tuncalp, 1988, 1999; Al-Meer, 1989; Abdul-Muhmin, 1998; Abdul-Muhmin and Alzame,2001). The two cities, Jeddah and Abha, werechosen since they are considered to be thetwo most attractive tourism cities in SaudiArabia. Jeddah is a coastal city and there are alot of entertainment activities. Abha city is asummer resort for Saudi Arabia and theweather is pleasant during the summer periodcompared to the majority of Saudi Arabia dueto the location and altitude. Out of 1400 ques-tionnaires distributed in Jeddah and Abha,505 usable questionnaires were verified andprepared for the final analysis.

Questionnaire

The 36 push items of motivation for pleasuretravel were rated on a five-point Likert scaleranging from 5 = strongly agree to 1 = stronglydisagree. For the push items, respondentswere requested to rate their level of agree-ment for each item when choosing a holidaytrip within Saudi Arabia. The 40 pull items ofattractiveness of a destination were rated onfive-point Likert scale ranging from 5 = veryimportant to 1 = very unimportant. For thepull items, respondents were requested to ratehow important each item was when selecting adestination for a domestic holiday trip.

Motivation for Domestic Tourism 53


Data Analysis and Discussion

In this research, the facility of the SPSS 10.0program to apply the alpha test wasemployed. The following results exhibit thescale reliability of the attitudinal statementsof tourist opinions on push and pull factors.The closer alpha is to 1 the more reliable theresults; push motivation with 36 items had analpha of 0.89 and pull motivation with 40items had an alpha of 0.86.

The analysis of data in this study consistedof two stages. First, factor analysis was employedon motivational items. Second, the correlationand regression analysis was used to examinethe nature of the relationship between pushand pull factors for pleasure tourism.

Table 5.1 presents the respondents’ demo-graphic characteristics. Seventy three per centof the respondents identified themselves asmale and 27% as female. They representedall age groups (25.7% under 25; 42.4%25–34; 22% 35–44 and 8.9% 45 or more).About 64.9% of the respondents were highlyeducated, with a university degree or above.They represented all monthly income levels(6.8% less than 3000 SR; 19.8% 3000–5999

SR; 28.3% 6000–8999 SR; 29.3% 9000–14,999SR; 15.8% 15,000 SR and more).

Factor Grouping of Push and PullMotivations Items

In order to identify motivation dimensions, 36push and 40 pull motivation items were sub-jected to a principal component factor analy-sis with Varimax orthogonal rotation usingSPSS factor analysis. Varimax rotation wasused to refine the original factors matrix.Rotation of the initial solutions maximizesvariance loading within each factor. Rotatingof the original matrix assists in the recogni-tion of the variables that best define the fac-tor. An orthogonal rotation was used for itssimplicity as suggested by Nunnally andBernstein (1994). A factor loading of 0.30 wasused as a cut-off for inclusion of any itemamong the various factors. The results of aprincipal component factor analysis withVarimax rotation produced a solution of ninepush and pull factors with eigenvalues greaterthan unity as shown in Tables 5.2 and 5.3.

These nine push factors explained 57.1%of the total variance and the nine pull factorsexplained 57.8% of the total variance. Hair etal. (1998) consider any solution with over60% of the explained variance (and in someinstance even less) as satisfactory from asocial sciences standpoint when informationis often less precise. The first factor for pushand pull explains the highest proportion ofobserved variance in the dataset. The secondfactor for push and pull accounts for themajority of variance not explained by factor 1,and so on. Factor loadings represent thedegree of correlation between an individualvariable and a given factor.

Reliability using Cronbach’s alpha was cal-culated to test the reliability and internal con-sistency of each push and pull factor. Theresults showed that the alpha coefficients forsix push factors ranged from 0.52 to 0.86.One factor, the relaxation factor, has slightlylower reliability (0.48). The interest factorand convenience of facilities factor had a lowreliability (0.44 and 0.33 respectively), but,the factor loadings were reasonably high, rep-resenting a respectable correlation between


Table 5.1. Description of survey respondents (n = 505).

Demographic variables %

SexMale 73Female 27

AgeLess than 25 year 25.725–34 year 42.435–44 year 22.045 year or more 8.9

Education levelPrimary or less 1.0Intermediate 3.8Secondary 30.3University 55.4Post-university 9.5

Monthly income levelLess than 3000 SR 6.83000–5999 SR 19.86000–8999 SR 28.39000–14,999 SR 29.315,000 SR and more 15.8



Tab

le 5

.2.

Fac

tor

load

ing

for

push

mot

ivat

ion-

rela

ted

item

s in

Sau

di to

uris

m.

Fac

tor

Var

ianc

eR

elia

bilit

ylo

adin

gsE

igen

valu

eex

plai

ned

Cro

nbac

h

Cul

tura

l val

ue8.

658

24.0

490.

86P

rese

rve

trad

ition

s an

d cu

stom

s w

hen

holid

ay in

Sau

di0.

779

Dom

estic

tour

ism

will

pro

tect

our

chi

ldre

n fr

om n

on-I

slam

ic v

alue

s0.

773

Hol

iday

s w

ithin

Sau

di a

re m

ore

appr

opria

te fo

r fa

mili

es th

an fo

reig

n ho

liday

s0.

772

The

soc

ial e

nviro

nmen

t will

hel

p w

omen

to p

rese

rve

the

Isla

mic

vei

l0.

764

Dom

estic

tour

ism

will

hel

p to

pro

tect

our

you

th fr

om s

uch

dise

ases

as

AID

S0.

728

It is

mor

e co

nven

ient

to p

erfo

rm Is

lam

ic r

itual

s w

hen

holid

ayin

g in

Sau

di th

an w

hen

holid

ayin

g in

no

n-Is

lam

ic c

ount

ry0.

714

Per

sona

l sec

urity

is a

ssur

ed w

hile

on

holid

ay in

Sau

di0.

681

To b

e w

ith o

ther

peo

ple

who

are

sim

ilar

to m

e in

thei

r tr

aditi

ons

and

cust

oms

0.65

8U

tilita

rian

2.50

06.

945

0.60

Spe

ndin

g a

holid

ay w

ithin

Sau

di w

ill h

elp

in o

btai

ning

new

exp

erie

nces

0.71

1N

o ne

ed to

cha

nge

Ara

bic

clot

hes

0.67

5B

ecau

se it

is c

heap

er th

an o

vers

eas

tour

ism

0.42

9K

now

ledg

e 1.

714

4.76

10.

58K

now

ing

how

oth

er p

eopl

e in

diff

eren

t reg

ions

live

0.73

4H

isto

rical

site

s ar

e ve

ry im

port

ant t

o m

y ho

liday

pla

ns0.

697

Get

ting

out i

nto

the

coun

try

and

rura

l are

as0.

619

Just

nat

ure

and

me,

that

’s m

y id

ea o

f per

fect

hol

iday

0.53

1S

ocia

l 1.

543

4.74

10.

67T

here

is fr

iend

ly tr

eatm

ent a

nd g

ener

osity

tow

ards

tour

ists

in S

audi

0.73

1C

oole

r w

eath

er r

esul

ts in

a m

ore

enjo

yabl

e ho

liday

0.65

7T

here

is a

goo

d se

lect

ion

of fa

mily

ent

erta

inm

ent p

lace

s0.

606

Dom

estic

hol

iday

s in

Sau

di A

rabi

a re

pres

ent g

ood

valu

e fo

r m

oney

0.34

0E

cono

mic

al1.

435

3.98

50.

61B

ecau

se h

olid

ay p

lace

s ar

e ne

ar to

my

hom

e th

ere

is n

o ne

ed to

pay

ext

ra a

ccom

mod

atio

n0.

669

To b

e ne

ar to

hol

y Is

lam

ic p

lace

s0.

510

Foo

d av

aila

bilit

y ac

cord

ing

to Is

lam

ic L

aw (

Sha

ria)

0.44

5F

amily

toge

ther

ness

1.

265

3.51

40.

52T

he p

erfe

ct h

olid

ay w

ould

incl

ude

all o

f our

fam

ily0.

667

The

yea

rly h

olid

ay is

the

time

whe

n th

e fa

mily

can

be

toge

ther

0.61

8Ta

lkin

g ab

out t

he p

lace

s I’v

e vi

site

d an

d th

e th

ings

I’ve

see

n0.

425

Aho

liday

in S

audi

Ara

bia

will

hel

p m

e re

st a

nd r

elax

0.31

3C

ontin

ued



Tab

le 5

.2.

Con

tinue

d.

Fac

tor

Var

ianc

eR

elia

bilit

ylo

adin

gsE

igen

valu

eex

plai

ned

Cro

nbac

h

Inte

rest

fact

or1.

210

3.36

10.

44D

oing

a lo

t of a

ctiv

ities

, lik

e sh

oppi

ng0.

766

Aho

liday

am

ong

peop

le is

ver

y en

joya

ble

0.63

8F

requ

ent s

hort

hol

iday

s of

fer

oppo

rtun

ity to

do

mor

e0.

430

Rel

axat

ion

1.13

23.

144

0.48

No

expe

nditu

re n

eede

d to

get

a v

isito

r vi

sa o

r to

cha

nge

curr

ency

0.56

7Ju

st to

cur

l up

with

a g

ood

book

in th

e sh

ade

soun

ds li

ke a

won

derf

ul h

olid

ay0.

537

Usu

ally

, we

visi

t rel

ativ

es o

r so

meo

ne w

e kn

ow o

n ou

r tr

ip0.

429

Hav

ing

fun,

bei

ng in

noce

ntly

ent

erta

ined

that

’s w

hat a

hol

iday

is a

ll ab

out

0.35

3F

rom

the

natio

nal e

cono

mic

poi

nt o

f vie

w s

pend

ing

mon

ey w

ithin

Sau

di w

ill b

e be

nefic

ial i

nste

ad o

f sp

endi

ng it

in a

noth

er c

ount

ry0.

353

Con

veni

ence

of f

acili

ties

1.09

53.

043

0.33

The

kin

ds o

f acc

omm

odat

ion

you

get a

re v

ery

impo

rtan

t0.

721

Ava

ilabi

lity

of g

ood

rest

aura

nts

and

good

food

is im

port

ant i

n a

holid

ay d

estin

atio

n0.

688



Tab

le 5

.3.

Fac

tor

load

ing

for

pull

mot

ivat

ion-

rela

ted

item

s in

Sau

di to

uris

m.

Fac

tor

Var

ianc

e R

elia

bilit

ylo

adin

gsE

igen

valu

eex

plai

ned

Cro

nbac

h

Saf

ety

6.72

216

.805

0.79

To fi

nd h

igh

stan

dard

s of

hyg

iene

/cle

anlin

ess

0.75

4To

feel

per

sona

lly s

afe

0.73

1To

feel

a w

arm

wel

com

e fo

r to

uris

ts0.

634

To e

njoy

the

cool

wea

ther

0.63

3To

find

qua

lity

rest

aura

nts

0.56

1To

be

easi

ly a

cces

sibl

e fo

r lo

cal a

ttrac

tions

0.55

7A

ctiv

ity3.

920

9.80

00.

78To

go

to fa

st fo

od r

esta

uran

ts0.

767

To g

o to

res

taur

ants

0.76

4To

enj

oy n

ight

life

0.66

5To

go

shop

ping

0.64

0To

go

to a

big

city

0.60

8B

each

spo

rts/

activ

ities

2.72

56.

813

0.79

To ta

ke p

art i

n w

ater

spo

rts

like

divi

ng a

nd w

ater

-ski

ing

0.72

1To

find

goo

d be

ache

s fo

r sw

imm

ing

0.69

1To

enj

oy p

ool a

ctiv

ities

0.65

3To

go

boat

ing

or to

cha

rter

a b

oat

0.64

7To

go

and

stay

at a

bea

ch c

abin

0.61

3To

go

fishi

ng0.

552

To v

isit

the

beac

h0.

515

Nat

ure/

outd

oor

2.17

95.

448

0.77

To p

artic

ipat

e in

out

door

act

iviti

es0.

732

To g

o to

a w

ilder

ness

are

as0.

715

To v

isit

mou

ntai

nous

are

as0.

682

To g

et a

way

from

cro

wds

0.65

1To

see

wild

life/

bird

s0.

613

His

toric

al/c

ultu

ral

1.96

84.

920

0.77

To v

isit

hist

oric

al a

rcha

eolo

gica

l site

s0.

685

To s

ee lo

cal c

rafts

/han

dwor

k0.

679

To m

eet i

nter

estin

g/fr

iend

ly lo

cal p

eopl

e0.

629

To v

isit

the

hist

oric

old

city

0.62

4To

vis

it in

tere

stin

g sm

all t

owns

/vill

ages

0.57

4To

see

out

stan

ding

sce

nery

0.46

4C

ontin

ued



Tab

le 5

.3.

Con

tinue

d.

Fac

tor

Var

ianc

e R

elia

bilit

ylo

adin

gs(a

)E

igen

valu

eex

plai

ned

Cro

nbac

h

Rel

igio

us1.

584

3.96

00.

86To

vis

it th

e K

aaba

0.87

4To

vis

it th

e P

roph

etic

Mos

que

0.87

1B

udge

t1.

546

3.86

60.

61To

find

bud

get a

ccom

mod

atio

n0.

684

To fi

nd v

alue

-for

-mon

ey r

esta

uran

ts0.

624

To fi

nd th

is tr

ip v

alue

-for

-mon

ey0.

620

To u

se p

ublic

tran

spor

tatio

n0.

480

Leis

ure

1.30

93.

273

0.64

To r

ide

tele

pher

age/

cabl

e ca

r0.

744

To v

isit

them

e pa

rk s

uch

as A

sir

Nat

iona

l Par

k0.

691

To g

o to

ent

erta

inm

ent/a

mus

emen

t pla

ces

0.44

9 U

psca

le1.

159

2.89

90.

63To

sta

y in

a fi

rst-

clas

s ho

tel

0.83

4To

sta

y in

a h

igh-

stan

dard

apa

rtm

ent/fl

at0.

759


items and the factor grouping to which theybelonged. Therefore, all nine factors wereincluded in the analysis; this is supported bythe work of Baloglu et al. (1998). Also, theresults showed that the alpha coefficients forthe nine pull factors ranged from 0.63 to0.79. The results are considered more thanreliable, since 0.50 is the minimum value foraccepting the reliability test (Nunnally, 1978).

The nine push factors are named ‘culturalvalue’, ‘utilitarian’, ‘knowledge’, ‘social’, ‘eco-nomical’, ‘family togetherness’, ‘interest’,‘relaxation’ and ‘convenience of facilities’.The nine pull factors accounted are named‘safety’, ‘activity’, ‘beach sports/activities’,‘nature/outdoor’, ‘historical/cultural’, ‘reli-gious’, ‘budget’, ‘leisure’ and ‘upscale’.

The mean is calculated to find the impor-tance for each push and pull factor as shown inTable 5.4. It can be seen that the cultural valuefactor is the most important push factor as per-ceived by Saudi tourists. The mean is 4.396,which means that domestic Saudi tourists like totake their holidays while at the same time pre-

serving their cultural values. The least impor-tant factor is the utilitarian factor with a meanof 3.595. It can be seen that the religious factoris the most important pull factor as perceived bySaudi tourists. The mean is 4.645, which meansthat the Holy Cities are the most attractive ofdestinations. The least important factor is thenature/outdoor factor with a mean of 3.513.

It can be concluded that the most impor-tant push factor is ‘cultural value’ and thisexplained 24.049% of the variance (the relia-bility coefficient is 0.86). The most importantpull factor is ‘religious’ and this explained16.805% of the variance (the reliability coeffi-cient is 0.79).

The evidence from this factor analysis sup-ports the push and pull motivations from pre-vious studies (Figler et al., 1992; Uysal andJurowski, 1994; Baloglu and Uysal 1996;Hanqin and Lam, 1999; You et al., 2000), inwhich they recognize that there are manymotivations that predispose the individuals totravel. However, the importance level of pushand pull factors might be different for tourists


Table 5.4. Importance ranking of push and pull motivation factors inSaudi tourism.

Motivation factors Mean Rank

Push factors Cultural value factor 4.396 1Convenience of facilities factor 4.359 2Family togetherness factor 4.164 3Social factor 4.044 4Knowledge factor 4.035 5Economical factor 3.941 6Interest factor 3.793 7Relaxation factor 3.763 8Utilitarian factor 3.595 9

Pull factorsReligious factor 4.645 1Safety factor 4.542 2Budget factor 3.892 3Leisure factor 3.878 4Upscale factor 3.717 5Historical/cultural factor 3.667 6Activity factor 3.551 7Beach sports/activities factor 3.546 8Nature/outdoor factor 3.513 9

Importance ranking were based on mean scores measured on theLikert five-point scale (5 = very important, 4 = important, 3 = neitherimportant or unimportant, 2 = unimportant, 1 = very unimportant).


from one country to another. In a mature mar-ket German tourists perceive ‘escape’ and‘active sports environment’ as the most impor-tant push and pull factors (Jamrozy and Uysal,1994). In addition, Australian tourists perceivethe most important push and pull factors as‘knowledge/intellectual’ and ‘historical/cul-tural’ (Oh et al., 1995). For a developing mar-ket, Mainland Chinese tourists perceive‘knowledge’ and ‘high-tech image’ as the mostimportant push and pull factors (Hanqin andLam, 1999). The top five push factors for theUK travellers were ‘going places I have not vis-ited before’, ‘being together as family’,‘increasing one’s knowledge about places’,‘people and things’, ‘visit friend and relatives’and ‘escaping from the ordinary’ (You et al.,2000). For Japanese travellers, the top push

factors were ‘going places I have not visitedbefore’, ‘having fun being entertained’, ‘get-ting a chance from busy job’, ‘just relaxing’and ‘increasing one’s knowledge about places,people and things’ (You et al., 2000).

This study found that the most importantpush and pull factors as perceived by Sauditourists were ‘cultural value’ and ‘religious’values.

Push and Pull Relationship

To examine the nature of the relationshipbetween push and pull factors for pleasuretourism the correlation and regression analy-sis was used. Table 5.5 shows the results ofcorrelation analysis among push and pull fac-


Table 5.5. Result of correlation and regression analysis of each push and each pull factor.

Pull factor

Regression coefficient

Beachsports/ Nature/ Historical/

Safety, Activity, activities, outdoor, cultural, Religious, Budget, Leisure, Upscale,adj. R2 adj. R2 adj. R2 adj. R2 adj. R2 adj. R2 adj. R2 adj. R2 adj. R2

Push factor = 0.18 = 0.23 = 0.08 = 0.12 = 0.36 = 0.06 = 0.04 = 0.12 = 0.01

Cultural value,adj. R2 = 0.30 0.28a NS �0.22a 0.15a 0.21a 0.22a �0.09b 0.23a �0.11b

Utilitarian,adj. R2 = 0.10 NS NS NS NS 0.22a NS NS 0.19a NSKnowledge,adj. R2 = 0.28 NS �0.12b NS 0.29a 0.40a NS NS NS NSSocial,adj. R2 = 0.11 0.13b 0.13b NS NS 0.20a NS NS 0.15a NSEconomical,adj. R2 = 0.15 �0.12b 0.13b 0.09c NS 0.21a 0.14a 0.17a 0.09c NSFamilytogetherness,adj. R2 = 0.04 NS 0.10c NS NS NS NS NS NS NSInterest,adj. R2 = 0.25 NS 0.43a 0.13b �0.16a 0.10c 0.09c NS NS 0.10c

Relaxation,adj. R2 = 0.08 NS NS 0.13b NS 0.12b NS 0.16a �0.12b NSConvenience offacilities,adj. R2 = 0.13 0.29a 0.15a NS NS NS NS NS NS 0.10c

aSignificant at the 0.001 level of significance.bSignificant at the 0.05 level of significance.cSignificant at the 0.01 level of significance.NS, not significant.


tors derived from factor analysis and regres-sion analysis. Each push factor was used as adependent variable to predict each one of thepull factors. Generally there exists some sig-nificant relationship between each push fac-tor and some of the pull factors at one ofthree levels of significance.

A regression model in which ‘the culturalvalue’ motivation was regressed on the ninepull factor revealed an adjusted R2 = 0.30, indi-cating that the model explained 30% of thetotal variance. Four pull factors – ‘safety’,‘nature/outdoor’, ‘historical/cultural’, ‘reli-gious’ and ‘leisure’ – were found to be impor-tant contributors in predicating ‘cultural value’motivations. A negative relationship betweenthe push factor ‘cultural value’ and the threepull factors ‘beach sports/activities’, ‘budget’and ‘upscale’ illustrates that such tourists didnot necessarily prefer these pull factors.

When ‘utilitarian’ is regressed on the ninepull factors (R2 = 0.10), less variance isexplained. The result of ‘utilitarian’ motiva-tion indicates that a significant relationship isfound between this factor and the two pullfactors ‘historical/cultural’ and ‘leisure’. Aregression equation to predicate ‘knowledge’motivations showed an adjusted R2 = 0.28,indicating that two pull factors ‘nature/out-door’ and ‘historical/cultural’, were signifi-cant. A negative relationship between thepush factor ‘knowledge’ and the pull factor‘activity’ was found.

When ‘social’ is regressed on the nine pullfactors (R2 = 0.11), less variance is explained.The finding of the ‘social’ push motivationsindicates that four pull factors ‘safety’, ‘activ-ity’, ‘historical/cultural’ and ‘leisure’, weresignificant. A regression in which ‘the eco-nomical’ motivation was regressed on the ninepull factors revealed an adjusted R2 = 0.15,indicating that the model explained 15% ofthe total variance. Five pull factors, ‘activity’,‘beach/sports activities’, ‘historical/cultural’,‘religious’ and ‘leisure’, were found to beimportant contributors in predicating ‘theeconomical’ motivations. A negative relation-ship between the push factor ‘economical’and two pull factor ‘safety’ and ‘leisure’ showsthat such tourists did not necessarily preferthese pull factors. A regression equation topredicate ‘the family togetherness’ motiva-

tions showed an adjusted R2 = 0.04, indicatingthat one pull factor, ‘activity’, was significantlylow. This is a factor in which it would beexpected for the family to take part.

‘Interest’ is regressed on the nine pull fac-tors (R2 = 0.25) indicating that the modelexplained 25% of the total variance. Five pullfactors, ‘activity’, ‘beach/sports activities’,‘historical/cultural’, ‘religious’ and ‘upscale’,were found to be important contributors inpredicating the ‘interest’ motivations. A nega-tive relationship between the push factor‘interest’ and one pull factor, ‘nature/out-door’, suggests that such tourists did not nec-essarily prefer this pull factor.

A regression equation to predicate thepush ‘relaxation’ factor showed an adjustedR2 = 0.08, indicating that the modelexplained 8% of the total variance. The resultindicates that three pull factors –‘beach/sports activities’, ‘historical/cultural’and ‘budget’ – were significant. A negativerelationship between the push factor ‘relax-ation’ and the pull factor ‘leisure’ was found.The ‘convenience of facilities’ is regressed onthe nine pull factors (R2 = 0.13) indicatingthat the model explained 13% of the totalvariance. Three pull factors – ‘safety’, ‘activ-ity’ and ‘upscale’ – were found to be impor-tant contributors in predicating ‘convenienceof facilities’ motivations.

It may be concluded that the results of thecorrelation analysis indicate that significantrelationships were found between the pushand the pull factor at the 0.001, 0.05 or 0.01level of significance. This supported previousresearch (Uysal and Jurowski, 1994; Balogluand Uysal, 1996; Kim and Lee, 2002) thatreported a relationship between push andpull factors.

Conclusions

This study provides the first attempt to exam-ine push and pull domestic tourism motiva-tion in Islamic and Arabic cultures, sincecultural variables play a significant role in thetourism motivation. Factor analysis producednine push motivational categories: ‘culturalvalue’, ‘utilitarian’, ‘knowledge’, ‘social’, eco-nomical’, ‘family togetherness’, ‘interest’,



‘relaxation’ and ‘convenience of facilities’.The nine pull motivational categories thatresulted from factor analysis are: ‘safety’,‘activity’, ‘beach sports/activities’, ‘nature/outdoor’, ‘historical/cultural’, ‘religious’,‘budget’, ‘leisure’ and ‘upscale’. By far themost important push and pull factors forSaudi tourists are ‘cultural value’ and ‘reli-gious’. There is predication in these factors, asthe Saudi culture is very much influenced byIslamic religion. The study also confirms therelationship between push and pull factors.

A limitation is that the sample populationof this study includes those visiting one oftwo tourist destinations, Jeddah and Abha.Although the sample size is not large, it islimited by its focus on only two tourist desti-nations, Jeddah and Abha. This gives rise toconcern that the findings based on a sampleof these towns might not generalize to alarger population of tourists. Also, applyinga convenience sample was necessitated bythe problems of sampling prevalent in theSaudi environment.


References

Abdul-Muhmin, A.G. (1998) Demographic differences in usage and attitudes toward the Saudi ArabianEFTPoS system. International Journal of Bank Marketing 16, 117–128.

Abdul-Muhmin, A.G. and Alzamel, I.A. (2001) Retailers’ experiences with and attitudes toward the SaudiArabian EFTPoS system. International Journal of Retail and Distribution Management 29, 188–199.

Al-Meer, A.A. (1989) Organizational commitments: a comparison of Westerners, Asians and Saudis.International Studies of Management and Organization 19, 74–82.

Al-Thagafy, S.A. (1991) Tourism patterns and behaviour in Saudi Arabia with special reference to theEmirate of Makkah Al-Mukarramah Province: a geographical study. PhD thesis, University ofSouthampton.

Asharq Alawsat Newspapers (1998) No. 7239, 23/9/1998.Baloglu, S. and Uysal, M. (1996) Market segments of push and pull motivation: a canonical correlation

approach. International Journal of Contemporary Hospitality Management 8, 32–38.Baloglu, S., Weaver, P. and McCleary, K.W. (1998) Overlapping product–benefit segments in the lodging

industry: a canonical correlation approach. International Journal of Contemporary Hospitality Management10, 159–166.

Beard, J.G. and Ragheb, M.G. (1983) Measuring leisure motivation. Journal of Leisure Research 15, 219–228.Cha, S., McCleary, K. and Uysal, M. (1995) Travel motivations of Japanese overseas travellers: a factor-clus-

ter segmentation approach. Journal of Travel Research 34, 33–39.Crompton, J. (1979) Motivation for pleasure vacation. Annals of Tourism Research 6, 408–424.Dann, G.M.S. (1977) Anomie, ego-enhancement and tourism. Annals of Tourism Research 4, 184–194.Dann, G.M.S. (1981) Tourist motivation: an appraisal. Annals of Tourism Research 8, 187–219.Economist Intelligence Unit Limited (1998) Welcome Mat. Crossborder Monitoring 6, 6.Figler, M.H., Weinstein, A.R., Sollers, J.J. III and Devan, B.D. (1992) Pleasure travel (tourist) motivation: a

factor analytic approach. Bulletin of Psychonomic Society 30, 113–116.Fodness, D. (1994) Measuring tourist motivation. Annals of Tourism Research 21, 555–581.Gilbert, D.C. (1991) An examination of the consumer behaviour process related to tourism. In: Cooper, C.P.

(ed.) Progress in Tourism, Recreation and Hospitality Management, 3. Belhaven Press, London, pp. 78–103.Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C. (1998) Multivariate Data Analysis, 5th edn. Prentice

Hall, Upper Saddle River, New Jersey. Hanqin, Z.Q. and Lam, T. (1999) An analysis of mainland Chinese visitors’ motivations to visit Hong Kong.

Tourism Management 20, 587–594.Holden, A. (1999) Understanding skiers’ motivations using Pearce’s ‘travel career’ construct. Annals of

Tourism Research 26, 435–438.Howells, B. (2000) Social and cultural impacts of tourism and the tourist motivation: Turkish Republic of

Northern Cyprus (TRNC) as a case study. In: Robison, M., Long, P., Evans, N., Sharpley, R. andSwarbooke, J. (eds) Motivations, Behaviour and Tourist Types. Centre for Travel and Tourism in associa-tion with Business Education Publishers, Sunderland, pp. 205–219.

Hudman, L.E. (1980) Tourism: a Shrinking World. Grid Publication, Columbus, Ohio.


Iso-Ahola, S.E. (1982) Toward a social psychological theory of tourism motivation: a rejoinder. Annals ofTourism Research 9, 256–262.

Jamrozy, U. and Uysal, M. (1994) Travel motivation variation of overseas German visitors. Journal ofInternational Consumer Marketing 6, 135–160.

Kim, S.S. and Lee, C.K. (2002) Push and pull relationships. Annals of Tourism Research 29, 257–260. Klenosky, D.B. (2000) The ‘pull’ of tourism destinations: a means-end investigation. Journal of Travel

Research 40, 385–395.Mansfeld, Y. (1992) From motivation to actual travel. Annals of Tourism Research 2, 399–419. Murphy, P.E. (1985) Tourism: a Community Approach. Methuen, New York.Nunnally, J.C. (1978) Psychometric Theory, 2nd edn. McGraw-Hall, New York.Nunnally, J.C. and Bernstein, I.H. (1994) Psychometric Theory, 3rd edn. McGraw-Hall, New York.Oh, C.H., Uysal, M. and Weaver, P.A. (1995) Product bundles and market segments based on travel motiva-

tions: a canonical correlation approach. International Journal of Hospitality Management 14, 123–137.Pearce, P.L. and Caltabiano, M.L. (1983) Inferring travel motivation from travellers’ experiences. Journal of

Travel Research 12, 16–20.Plog, S.C. (1974) Why destination areas rise and fall in popularity. Cornell Hotel and Restaurant Quarterly 14,

55–58.Sajini, I. (1997) Numbers in tourism. Al-Madinah Newspaper, No.12468, 3.6.Sirakaya, E. and McLelland, R.W. (1997) Factors affecting vacation destination choices of college students.

Anatolla: An International Journal of Tourism and Hospitality Research 8, 31–44.Sharpley, R. (1999) Tourism, Tourists and Society, 2nd edn. ELM, Huntingdon.Tuncalp, S. (1988) The marketing research scene in Saudi Arabia. European Journal of Marketing 22, 15–22.Tuncalp, S. (1999) Evaluation of information sources in industrial marketing: implications for media plan-

ning in the Arabian Gulf. Journal of Business & Industrial Marketing 14, 49–60.Uysal, M. and Hagan L.A.R. (1993) Motivation of pleasure travel and tourism. In: Khan, M., Olsen, O. and

Var, T. (eds) VNR’s Encyclopaedia of Hospitality and Tourism. Van Nostrand Reinhold, New York, pp.798–810.

Uysal, M. and Jurowski, C. (1994) Testing the push and pull factors. Annals of Tourism Research 21, 844–846.Yavas, U., Cavusgil, S.T. and Tuncalp, S. (1987) Assessments of selected foreign suppliers by Saudi

importers: implications for expatriate. Journal of Business Research 15, 137–246.You, X., O’Leary, J., Morrison, A. and Hong, G.S. (2000) A cross-cultural comparison of travel push and

pull factors: United Kingdom vs. Japan. International Journal of Hospitality & Tourism Administration 1,1–26.



Chapter sixEcotourists’ Environmental Learning Opportunity as aSource of Competitive Advantage: Are Ecotourism OperatorsMissing the Boat with their Advertising?

Garry G. PriceSchool of Tourism and Hospitality, La Trobe University, Bundoora, Victoria 3086, Australia

Abstract

Research indicates that tourist motivation and behaviour are often determined, in part, by the desire fora learning experience. It is contended in this chapter that ecotourists have a desire for environmentallearning and that ecotourism operators who differentiate their product through the provision of high-quality environmental education programmes establish a potential sustainable competitive advantage.However, it is asserted that this potential long-term competitive advantage will only be realized if suffi-cient tourists can be attracted to the offered ecotourism experience. It is further asserted that attractingthe required number of clients can be done, in part, by ecotourism operators emphasizing environmentallearning in their printed and electronic advertising material. Content analysis of the current advertisingmaterial of ecotourism operators reveals that many ecotourism operators are not placing emphasis ontheir environmental learning offerings in their advertising. Using the results obtained, it is argued thatthe relatively small emphasis on ecotourists’ demand for environmental learning in advertising directedat potential clients is a substantial weakness in some companies’ strategic marketing. It is further arguedthat authentic and expert ecotourism operators who capitalize on this shortcoming are likely to achieve asustainable competitive advantage.

Introduction

This chapter is based on the contentions thatecotourists have a desire for environmentallearning and that ecotourism operators whodifferentiate their product through the provi-sion of high-quality environmental educationprogrammes establish a potential sustainablecompetitive advantage. However, it is also

contended that this potential long-term com-petitive advantage will only be realized if suffi-cient tourists can be attracted to the offeredecotourism experience. This can be done, inpart, by ecotourism operators emphasizingthe provision of environmental learningexperiences in their printed and electronicadvertising material. The degree to which thisis done is the focus of this chapter.



The Target Market

Despite many attempts to develop a precise,universally accepted definition of ecotourism,and hence an ecotourist (Ceballos-Lascurain,1988; Romeril, 1989; Scace et al., 1992; Figgis,1993; Richardson, 1993; Valentine, 1993;Western, 1993; Fennell, 1999; Weaver, 2001),the definition remains problematic. However,one definition worthy of explicit inclusion inthis paper, because of its generally acceptedconceptual content, is that of the AustralianNational Ecotourism Strategy which statesthat:

Ecotourism is nature-based tourism that involveseducation and interpretation of the natural envi-ronment and is managed to be ecologically sus-tainable. The definition recognizes that ‘naturalenvironment’ includes cultural components andthat ‘ecologically sustainable’ involves an appro-priate return to the local community and long-term conservation of the resource.

(Commonwealth Department of Tourism, 1994, p. 17)

Wood and House (1991) took account ofsome aspects of ecotourism when they classi-fied tourists according to their behaviourtowards the environment and the host com-munity in their holiday destination. Theassertion was made that tourists could be clas-sified as ‘good’ tourists if their behaviour, atleast in part, embraces the principles of sus-tainability. That is, the tourists’ behaviourreflects the need to safeguard and enhancethe natural and cultural assets of the destina-tion, to safeguard and enhance the residentpopulation’s quality of life and life opportuni-ties, and to allow a return on investment fortourism operators (Faulkner, 2001).

Some of these ‘good’ tourists could, ofcourse, be classified as ecotourists if their cho-sen destinations and activities meet the gener-ally accepted ecotourism principles of beingnature-based, ecologically, economically andsocio-culturally sustainable, environmentallyeducative and locally beneficial (Price andMurphy, 2000). This demand-side categoriza-tion of ecotourists is somewhat appealingwith its emphasis on accepting nature on itsown terms. However, as discussed below,some tourists who might be classified as psy-chocentric (Plog, 1973), and who are there-

fore excluded from the generally acceptedcategorization of ecotourists (Weaver, 2001,p. 47), participate in ecotourism activities(Jackson et al., 2000). This observation makesa purely demand-side categorization of eco-tourists invalid.

The assertion by Sirakaya et al. (1999) thatecotourism, and hence ecotourists, should bedefined from the supply side, dependent on anoperator’s evaluation of the experience, shouldalso be treated with caution. The term ‘eco-tourism’ sometimes has been used as a market-ing ploy or political tool (Rees, 1990; Wight,1993; Cater and Lowman, 1994) and, conse-quently, supply-side definitions and evaluationsshould be viewed warily. As has been claimed:

Ideally, [ecotourism] companies will demon-strate sensitivity to environmental concerns, tryto operate within the framework of ecologicalguidelines, promote an ecologically correctawareness, and support/use local tourism ser-vices … practice [sic] longterm planning, andcontinue to support destination communities’conservation/preservation programs.

(Sirakaya et al., 1999, p. 169; author’s emphasis)

It is obvious from the above discussion that itis very difficult, if not impossible, to develop aprecise definition of an ecotourist.Consequently, in this chapter ecotourists arecategorized as clients of companies whichpurport to be ecotourism businesses. This cat-egorization does not overcome the problemof definitional precision but it does provide afocus for the chapter. It also recognizes theinterdependence of supply and demand inecotourism (Uysal, 2000). Supply-side factors,such as the development and provision ofenvironmental education programmesdepend, at least in part, on the demands ofecotourism clients who want environmentallearning to be part of their tourism experi-ence. This recognition is seen by some acade-mics as being critical to comprehensiveecotourism analyses (Burgess, 1991; Nelson,1994; Markwell, 1998; Sun and Walsh, 1998).

Research indicates that people who mightbe categorized as ecotourists come fromdiverse demographic and lifestyle backgrounds.However, despite this diversity, it has beendetermined that the majority of ecotourists arewell educated, have above-average incomes and

66 G.G. Price


possess interest in the natural environment,although to varying degrees (Fennell andSmale, 1992; Hvenegaard, 1994; Eagles andCascagnette, 1995; Blamey, 1997; Beeton,1998).

Notwithstanding the difficulty of ascertain-ing precisely who is an ecotourist, there hasbeen some consideration of the reasons fortourists choosing experiences which fall into thebroad ecotourism classification. Much of theacademic discourse of tourist motivation hasarisen from the adaptation of needs theories,especially Maslow’s (1943) hierarchy of needstheory, to tourism motivation (Decrop, 1999).Krippendorf (1987) provided an overall sum-mary of the various theories by stating thattravel is motivated by escaping from somethingrather than going to something and that tourists’motives are primarily self-orientated, a view inaccord with Huxley’s (1925, p. 12) reflectionthat: ‘We read and travel not that we maybroaden our minds, but that we may pleasantlyforget they exist’. However, it has become evi-dent in recent years that some tourists, at least,are travelling for specific purposes includingthe enrichment of their lives (MacKay, 1995), apositive response not based on escapism.According to Holbrook and Hirschman (1982)tourists’ decisions can be based on emotionalarousal. This arousal is sometimes facilitated bythe qualities of a particular destination (Miles,1991; Uysal and Hagan, 1993; Prentice et al.,1998) and by the need for a learning experi-ence (Swarbrooke and Horner, 1999, p. 60).

Plog (1973) indicated that there was adirect link between the psychographic classifi-cation of tourists and their destinations andactivities, an assertion that has been justifiablyquestioned over the past few years (Lee-Hoxter and Lester, 1988; Smith, 1990;Nickerson and Ellis, 1991; Jackson et al.,2000). In recent research, Jackson et al.(2000) reported that, although there was astatistically significant relationship betweenpersonality type and motivation for a preferredactivity, actual choice is likely to be influencedby subjective norms (the approval of others)and the perceived control over the decision(contextual factors).

Pizam and Mansfield (1999) asserted thatmotivation is but one of the factors that influ-ence consumer behaviour and that other fac-

tors such as perception, learning, personalityand attitudes should be taken into account.Baloglu and Uysal (1996, p. 32) argued simi-larly, but identified the relative importance ofmotivation, asserting that ‘although motiva-tion is only one variable explaining touristbehaviour, it is regarded as one of the mostimportant variables because it is an impellingand compelling force behind all behaviour’.

Eagles (1991) showed that Canadian eco-tourists (defined in his study as people whohad participated in specific ecotourism activi-ties) have motivations different from othertravellers. Among ecotourists’ motivationsemphasis was placed on a desire for natureand/or wilderness experiences. However,other non-nature motivations such as thedesire to observe simpler lifestyles and thedesire to rediscover self were evident, resultssupported by Woods and Moscardo (1998).These findings also were consistent with thoseof Silverberg et al. (1996) who found that eco-tourists have a variety of needs, wants andmotives for their travel behaviour. Amongthese needs, wants and motives is learningabout the environment (Swarbrooke andHorner, 1999, p. 60).

What is Environmental Learning?

Theorists, researchers and practitioners haveproduced many and varied definitions oflearning and there has been considerable dis-agreement over the precise nature of learn-ing (Shuell, 1986; Malone, 1991; Schunk,1996; Phillips and Soltis, 1998). However,most definitions incorporate the notion thatlearning involves the acquisition and modifi-cation of knowledge, skills, strategies, beliefs,attitudes and behaviours.

One definition that captures the criteriamost scholars consider central to learningstates that:

Learning is an enduring change in behaviour,or in the capacity to behave in a given fashion,which results from practice or other forms ofexperience.

(Shuell, 1986)

There is still no universally accepted succinctdefinition of environmental learning, or theadvancement of environmental literacy. How-

Environmental Learning as a Competitive Advantage 67


ever, it is generally accepted that environmen-tal literacy involves an increase in awarenessof, and concern about, economic, social,political, and ecological interdependence inurban and rural areas. This increase is depen-dent on the acquisition of the knowledge, val-ues, attitudes, commitment and skills neededto protect and improve the environment andto create new patterns of behaviour towardsthe environment (Tbilisi Declaration, 1978).

This broad statement on environmentalliteracy has been clarified and refined overthe years (Roth, 1991; Fien, 1993a–d, 1996;Payne, 1995). Roth (1991), for example,asserted that individuals progress along a con-tinuum towards the acquisition of environ-mental literacy, namely:

1. The nominal stage, in which an individualhas some knowledge of basic environmentalterms but has little depth of understandingand has no more than a casual commitmentto environmental concerns and actions.2. The functional stage, in which an individ-ual has the capacity to use fundamental envi-ronmental knowledge to formulate actionpositions on particular environmental issuesand in daily behaviour.3. The operational stage, in which an individ-ual has the capacity to regularly perceive envi-ronmental issues, gather and evaluaterelevant information, examine and chooseamong alternatives and take positions andactions that work to sustain and develop thefoundations of environmental knowledge.

Movement along the EnvironmentalLiteracy Continuum

At the heart of environmental learning, orthe development of environmental literacy,within an ecotourism context lies the conceptof interpretation which was first defined byTilden (1977, p. 9) as:

An educational activity which aims to revealmeanings and relationships through the use oforiginal objects, by firsthand experience, and byillustrative media, rather than simply to commu-nicate factual information.

According to the Nature and EcotourismAccreditation Program (2000, p. 130), interpre-

tation ‘is the art of bringing together manypieces of information and relating them tothe setting or experience in such a way that itall becomes more meaningful and enjoyable’.Also within recent interpretation, consider-able emphasis has been placed on the devel-opment of affection for the environment‘with the hope that such affection will lead tocareful treatment’ (Kimmel, 1999, p. 42).This view supports the emphasis by Ham(1992, pp. 33–43) on a thematic (principalmessage) approach, rather than a topic (sub-ject matter) approach, to environmentallearning in informal settings.

One possible motivating factor of particu-lar interest to this author is that of ecotouristsextending their affection for the environ-ment to seek a deeper and more meaningfulconnection with the natural world. This moti-vational factor can be described as numen-seeking1, the term ‘numen’ being originallyattributed to Otto (1958), and used by him todescribe a religious emotion or experience inthe presence of something holy. The termsubsequently has been used in varioushumanities to describe experiences, placesand objects which inspire reactions of rever-ence and awe (Maines and Glynn, 1993).Oubre (1997, p. 215), for example, in dis-cussing the evolution of human conscious-ness, sees the numinous mind as allowingtranscendental thought which she defines as:

direct experiential apprehension of perceptionand consciousness outside the realm of normalhuman experience that is said to bring serenity,peace, understanding and deep knowledge.

Prentice et al. (1998) implied that significantheritage sites can induce numinous reactionsin visitors, a view that academics such as Lewis(1985) and Miles (1991) have applied to thenatural environment. In emphasizing theemotional aspects of wilderness learning, forexample, Miles (1991, p. 6) asserted that:

Learning about wilderness is not like learningarithmetic or economics or how the politicalsystem works. It is more akin to learning abouta Mozart concerto, a Rembrandt painting, or aShakespearean sonnet.

It is contended in this chapter that eco-tourism operators who offer the setting toattract ecotourists, and who enable them to

68 G.G. Price


progress along the continuum of environ-mental literacy through the provision ofmeaningful and enjoyable environmentallearning experiences, including the satisfyingof numinous desires, have a potential long-term competitive advantage.

Environmental Learning and SustainableCompetitive Advantage

According to Lewis et al. (1993, p. 20)

only two forms of competitive advantage exist:1. Cost leadership – lower total relative costs;2. Differentiation – unique benefits that attract aprice premium.

It is the latter item that is the focus of thischapter. The main assertion of the author isthat the provision of informal environmentaleducation by ecotourism operators differenti-ates ecotourism from other forms of nature-based tourism (The Economic PlanningGroup of Canada, 2000, p. 6) and hence canprovide a sustainable competitive advantage tothose operators who embrace environmentaleducation enthusiastically and responsibly.This assertion is based on the argument ofBarney (1986, cited in Lewis et al., 1993,28–29) that to be a source of sustainable com-petitive advantage a firm’s capabilities must bevaluable, rare and hard to imitate. Within eco-tourism, some operators possess attributessuch as a high level of ecological knowledgeand highly developed interpretive programmefacilitation skills that are valuable to clients,not possessed by other nature-based tourismoperators and are difficult to emulate. Suchattributes, which can contribute to the provi-sion of high-quality informal environmentaleducation, have the potential to provide a sus-tainable competitive advantage to an opera-tor’s company. However, this potentialsustainable competitive advantage will only berealized if sufficient clients can be attracted tothe ecotourism experience. This attraction is,in part, dependent on the content of the pro-motional material, both printed and elec-tronic, of the firm conducting the ecotourismoperation. Content analysis of such materialdoes not appear to have been carried out inany objective and systematic way.

Content Analysis

Content analysis has been defined in a num-ber of ways (Holsti, 1969; Carney, 1972;Weber, 1985). Despite the range of viewsexpressed, the definitions reveal broad agree-ment on the requirements of an objectiveand systematic approach to content analysisand the need for consequent comparisonbetween content data and some other data.According to Holsti (1969, pp. 3–4) objectiv-ity is achieved through the application ofexplicitly formulated rules and proceduresthat minimize the subjective predispositionsof the researcher. System is attained throughthe inclusion and exclusion of contentaccording to consistently applied rules so thatanalyses based only on material supportingthe researcher’s hypotheses are eliminated.In this chapter, content analysis is taken to be‘any technique for making inferences byobjectively and systematically identifying spec-ified characteristics of messages’ (Carney,1972, p. 25).

Content Analysis of Advertising Materialof Ecotourism Companies

Source of data

Students studying the third-year subjectTourism and the Environment at La TrobeUniversity, Victoria, Australia, in 2002 (78 stu-dents) collected 254 examples of the promo-tional material of companies whichpurported to be ecotourism companies. Thematerial was obtained from both electronicand printed sources, and the purpose of thecollection was not revealed to the studentsuntil the collection was completed.

Method

Content analysis of the collected material wasperformed. When this analysis of the promo-tional material of the ecotourism companieswas carried out, objectivity was achieved byhaving three investigators (two students andthe author), all of whom had detailed knowl-edge of the subject matter, independently



apply content analysis to each article underconsideration. A systematic approach to avoidbias in the selection of material was accom-plished by the collection of material by per-sons who had no knowledge of theinvestigative thrust of this author.

Since the focus of this chapter is the envi-ronmental education component of eco-tourism, only this key principle ofecotourism was considered. Identifyingwords within the category were education,learning, knowledge, attitude, behaviour,teaching, facilitation and interpretation. Thenumber of the words in the sentences relat-ing to the category allowed the proportionalemphasis on environmental learning to bedetermined, assuming that higher relativecategory counts reflect higher concern(Weber, 1985, p. 56). The data obtainedquantitatively were rejected if there was anumerical discrepancy of greater than 10%between the determined values of the inves-tigators. This rejection is in accord with thesuggestion of Holsti (1969, pp. 3–4) to applyexplicitly formulated rules and proceduresto minimize subjectivity. The mean percent-age of the three acceptable values for eachitem was used to obtain the results.

Results

The content analysis performed on the 117valid items revealed that environmental edu-cation is relatively neglected in the selectedecotourism companies’ promotional material.The mean value of the percentage of materialdevoted to the environmental education com-ponent of ecotourism was found to be 13.5%with a standard deviation of 13.31, as shownin Fig. 6.1.

Discussion and Conclusion

The results obtained from this research indi-cate that many ecotourism operators are notemphasizing their environmental learningofferings in their promotional material. Itshould be noted, however, that the approachto content analysis followed in this chapterprovides data only at the time of judgement.Given that there is a relatively fixed amountof advertising material produced by eachcompany, further analysis over time wouldenable judgement of the relative increase ordecrease in emphasis on environmentallearning. However, this was not feasible given

70 G.G. Price

Fre

quen

cy

30

20

10

00.0

5.010.0

15.020.0

25.030.0

35.040.0

45.050.0

55.060.0

Environmental education (%)

SD = 13.31Mean = 13.5n = 117.00

Fig. 6.1. Histogram of the percentage of ecotourism operators’ advertising devoted to environmentaleducation.


the time constraints of this work.Consequently, the analysis conducted pro-vides a ‘snapshot’ of the current environmen-tal learning emphasis of each of the selectedcompanies and helps to establish an industrynorm. This inductive approach recognizesthat the development of some a priori stan-dard of adequacy or performance is not validin the context of this chapter, but that anindustry norm against which the advertisingmaterial of any single operator can be com-pared is able to be obtained.

Notwithstanding this comparison, it is rec-ognized by this author that, however objectiveand systematic the measures of communica-tion content, inferences about the actual eco-tourism operation cannot legitimately bemade from content data alone.Corroborating evidence from independent,non-content data must also be used and thisfocus is contained in the author’s furtherresearch to determine the effectiveness ofecotourism companies’ promotional material

in attracting clients. Given the findings ofother researchers, however, that ecotouristsare seeking environmental learning experi-ences, the fact that there is only a relativelysmall emphasis on this demand in the adver-tising directed at potential clients is a substan-tial weakness in some companies’ strategicmarketing. It is contended that authentic andskilful ecotourism operators who capitalize onthis advertising shortcoming are likely toachieve a sustainable competitive advantageby attracting a greater number of clients totheir operation and by satisfying their clients’needs in ways which are hard to emulate.

Note

1The author would prefer to use the Platonian term‘epithumetic motivation’ rather than ‘numen seek-ing’ because of the religious origins of the latterterm. However, because of the current relativelywidespread use of ‘numen seeking’ in non-religiouscontexts, the term has been used in this chapter.


References

Baloglu, S. and Uysal, M. (1996) Market segments of push and pull motivations: a canonical correlationapproach. International Journal of Contemporary Hospitality 8, 32–38.

Beeton, S. (1998) Ecotourism: a Practical Guide for Rural Communities. Landlinks Press, Collingwood, Victoria.Blamey, R.K. (1997) Ecotourism: the search for an operational definition. Journal of Sustainable Tourism 5,

109–130.Burgess, J. (1991) Softly minimising the impact of ecotourism in Tasmania. In: Weiler, B. (ed.) Ecotourism:

Incorporating the Global Classroom. Bureau of Tourism Research, Canberra, pp. 89–93.Carney, T.F. (1972) Content Analysis: a Technique for Systematic Inference from Communications. B.T. Batsford,

London.Cater, E. and Lowman, G. (eds) (1994) Ecotourism: a Sustainable Option? John Wiley & Sons, Chichester.Ceballos-Lascurain, H. (1988) The future of ‘ecotourism’. Mexico Journal 27, 13–14. Commonwealth Department of Tourism (1994) National Ecotourism Strategy. Australian Government

Publishing Service, Canberra.Decrop, A. (1999) Tourists’ decision-making and behavior process. In: Pizam, A. and Mansfield, Y. (eds)

Consumer Behavior in Travel and Tourism. Haworth Press, New York.Eagles, P.F. (1991) The motivations of Canadian ecotourists. In: Weiler, B. (ed.) Ecotourism: Incorporating the

Global Classroom. Bureau of Tourism Research, Canberra.Eagles, P.F. and Cascagnette, J. (1995) Canadian ecotourists: who are they? Tourism Recreation Research 20,

22–28.Ecotourism Association of Australia (2000) Nature and Ecotourism Accreditation Program, 2nd Edition.

Ecotourism Association of Australia, Brisbane.Faulkner, B. (2001) Destination Australia: a research agenda for 2002 and beyond. Paper presented at

Australian Tourism Research Institute Conference, 30–31 October 2001, Hobart.Fennell, D.A. (1999) Ecotourism: an Introduction. Routledge, London.Fennell, D.A. and Smale, B.J.A. (1992) Ecotourism and natural resource protection: implications of an

alternative form of tourism for host nations. Tourism Recreation Research 17, 21–32. Fien, J. (1993a) Educating for the Environment: Critical Curriculum Theorising and Environmental Education.

Deakin University Press, Geelong, Victoria.


Fien, J. (ed.) (1993b) Environmental Education: a Pathway to Sustainability. Deakin University Press, Geelong,Victoria.

Fien, J. (1993c) Teaching for a sustainable world: a new agenda in teacher education. In: Fien, J. (ed.)Teaching for a Sustainable World: Environmental and Development Education Project for Teacher Education.Australian Association for Environmental Education, Brisbane.

Fien, J. (1993d) Introduction to environmental education. In: Fien, J. (ed.) Teaching for a Sustainable World:Environmental and Development Education Project for Teacher Education. Australian Association forEnvironmental Education, Brisbane.

Fien, J. (1996) Learning for a Substainable Future. UNESCO, Paris.Figgis, P.J. (1993) Ecotourism: special interest or major direction? Habitat 21, 8–10. Ham, S.H. (1992) Environmental Interpretation: a Practical Guide for People with Big Ideas and Small Budgets.

North American Press, Golden, Colorado.Holbrook, M.B. and Hirschman, E.C. (1982) The experiential aspects of consumption: consumer fantasies,

feelings and fun. Journal of Consumer Research 23, 132–140.Holsti, O.R. (1969) Content Analysis for the Social Sciences and Humanities. Addison-Wesley, Reading,

Massachusetts.Huxley, A. (1925) Along the Road. Chatto and Windus, London.Hvenegaard, G.T. (1994) Ecotourism: a status report and conceptual framework. Journal of Tourism Studies

5, 24–35.Jackson, M.S., White, G.N. and Schmierer, C.L. (2000) Predicting tourism destination choices: psycho-

graphic parameters versus psychological motivations. In: Michael, E. (ed.) Peak Performance in Tourismand Hospitality Research CAUTHE 2000 Refereed Research Papers. School of Tourism and Hospitality, LaTrobe University, Bundoora, pp. 57–63.

Kimmel, J.R. (1999) Ecotourism as environmental learning. The Journal of Environmental Education 30,40–45.

Krippendorf, J. (1987) The Holidaymakers. Heinemann, London.Lee-Hoxter, A. and Lester, D. (1988) Tourist behaviour and personality. Personality and Individual Differences

9, 177–178.Lewis, G., Morkel, A. and Hubbard, G. (1993) Australian Strategic Management: Concepts, Context and Cases.

Prentice Hall, Sydney.Lewis, P. (1985) Presidential address: beyond description. Annals of the Association of American Geographers

75, 467.MacKay, M.-J. (1995) Gauging the impact of opportunity: an analysis of sustainable ecotourism in Cape

Breton native communities. [Web page] Available: http://fortress.uccb.ns.ca/search/Proc95_4.htm[accessed 5 February, 2002].

Maines, R. and Glynn, J. (1993) Numinous objects. The Public Historian 15, 9–25.Malone, J.C. (1991) Theories of Learning: a Historical Approach. Wadsworth Publishing Company, Belmont,

California.Markwell, K. (1998) Ecotourism: nature protected or nature commodified? In: Rowe, D. and Lawrence, G.

(eds) Tourism, Leisure, Sport: Critical Perspectives. Hodder Education, Rydalmere, New South Wales,pp. 63–73.

Maslow, A.H. (1943) A theory of human motivation. Psychological Review 50, 370–396.Miles, J. (1991) Viewpoint: teaching in wilderness. The Journal of Environmental Education 22, 5–9.Nelson, J.G. (1994) The spread of ecotourism: some planning implications. Environmental Conservation 21,

248–255.Nickerson, N.P. and Ellis, G.D. (1991) Traveller type and activation theory: a comparison of two models.

Journal of Travel Research 29, 26–31.Otto, R. (1958) The Idea of the Holy. Oxford University Press, London.Oubre, A.Y. (1997) Instinct and Revelation: Reflections on the Origins of Numinous Perception. Gordon and

Breach, Amsterdam.Payne, P. (1995) Ontology and the critical discourse of educational education. Australian Journal of

Environmental Education 11, 83–105.Phillips, D.C. and Soltis, J.F. (1998) Perspectives on Learning, 3rd edn. Teachers College Press, New York. Pizam, A. and Mansfield, Y. (eds) (1999) Consumer Behavior in Travel and Tourism. Haworth Press, New

York.Plog, S.C. (1973) Why destination areas rise and fall in popularity. Cornell Hotel and Restaurant

Administration Quarterly 13, 13–16.

72 G.G. Price


Prentice, R., Witt, S. and Hamer, C. (1998) Tourism as experience: the case of heritage parks. Annals ofTourism Research 25, 1–25.

Price, G.G. and Murphy, P.E. (2000) The relationship between ecotourism and sustainable development: acritical examination. In: Michael, E. (ed.) Peak Performance in Tourism and Hospitality Research CAUTHE2000 Refereed Research Papers. School of Tourism and Hospitality, La Trobe University, Bundoora,pp. 188–201.

Rees, N.E. (1990) The ecology of sustainable development. The Ecologist 20, 18–23.Richardson, J.A. (1993) Ecotourism and Nature-Based Holidays. Choice Books, Marrickville, New South Wales.Romeril, M. (1989) Tourism and the environment: accord or discord? Tourism Management 10, 204–208.Roth, C.E. (1991) Towards shaping environmental literacy for a sustainable future. ASTM Standardization

News (April), 42–45.Scace, R.C., Grifone, E. and Usher, R. (1992) Ecotourism in Canada. Ministry of Supply and Services, Hull,

Quebec.Shuell, T.J. (1986) Cognitive conceptions of learning. Review of Educational Research 56, 411–436.Silverberg, K.E., Blackman, S.J. and Backman, K.F. (1996) A preliminary investigation into the psycho-

graphics of nature-based travelers to the southeastern United States. Journal of Travel Research 35,19–28.

Sirakaya, E., Sasidharan, V. and Sonmez, S. (1999) Redefining ecotourism: the need for a supply-side view.Journal of Travel Research 38, 168–172.

Smith, S.L.J. (1990) A test of Plog’s allocentric/psychocentric model: evidence from seven nations. Journalof Travel Research 28, 40–43.

Sun, D. and Walsh, D. (1998) Review of studies on environmental impacts of recreation and tourism inAustralia. Journal of Environmental Management 53, 323–338.

Swarbrooke, J. and Horner, S. (1999) Consumer Behaviour in Tourism. Butterworth-Heinemann, Oxford.Tbilisi Declaration (1978) Connect 3, 1–8. The Economic Planning Group of Canada (2000) On the path to success: Canadian ecotourism [Web

page] Available http://www.canadatourism.com/en/ctc/partner_center/partnering/ind_comp.cfm[Accessed 6 May, 2000].

Tilden, F. (1977) Interpreting our Heritage, 3rd edn. University of North Carolina Press, Chapel Hill.Uysal, M. (2000) Keynote address. Peak Performance in Tourism and Hospitality Research CAUTHE 2000.

School of Tourism and Hospitality, La Trobe University, Bundoora.Uysal, M. and Hagan, L.A.R. (1993) Motivation of pleasure travel and tourism. In: Kham, M.A., Olsen,

M.D. and Turgut, V. (eds) Encyclopaedia of Hospitality and Tourism. Van Nostrand Reinhold, New York,pp. 798–821.

Valentine, P.S. (1993) Ecotourism and nature conservation: a definition with some recent developments inMicronesia. Tourism Management 14, 107–115.

Weaver, D. (2001) Ecotourism. John Wiley & Sons, Brisbane.Weber, R.P. (1985) Basic Content Analysis. Sage Publications, Beverly Hills, California.Western, D. (1993) Defining ecotourism. In: Lindberg, K. and Hawkins, D. (eds) Ecotourism: a Guide for

Planners and Managers. The Ecotourism Society, North Bennington, Vermont, pp. 7–11. Wight, P.A. (1993) Ecotourism: ethics or eco-sell? Journal of Travel Research 31, 3–9.Wood, K. and House, S. (1991) The Good Tourist: a Worldwide Guide for the Green Traveller. Mandarin,

London.Woods, B. and Moscardo, G. (1998) Understanding Australian, Japanese and Taiwanese ecotourists in the

Pacific Rim region. Pacific Tourism Review, 329–339.



Chapter sevenDomestic Leisure Traveller Consumption Systems

Elizabeth Cowley,1 Ray Spurr,1 Peter Robins2 and Arch G. Woodside3

1School of Marketing, University of New South Wales, Sydney, Australia; 2Bureau of TourismResearch, GPO Box 1545, Canberra, ACT, 2601 Australia; 3Carroll School of Management,Boston College, 450 Fulton Hall, 140 Commonwealth Avenue, Chestnut Hill, MA 02467, USA.

Abstract

The planning of tourism marketing strategies probably improves by focusing on specific visitor consumptionsystems, and avoiding attempts to promote products–services in consumption systems unrelated to the mar-keter’s destination or service. Using unit record data of overnight domestic travellers (n = 27,653 households)in the Australian National Visitor Survey (NVS) 1998 study, this chapter reports findings of an empirical studythat support the following core propositions. Leisure travellers visiting overnight destinations complete a seriesof related purchases that may be viewed usefully as a leisure travel consumption system. Certain travel-relatedpurchases in one product–service category trigger additional purchases in other product–service categories.Substantial numbers of leisure visitors to each major destination (i.e. city, region, attraction or event) can besegmented usefully into micro-segments according to their purchase/consumption systems. For example, win-ter or summer Australian domestic visitors to Brisbane, Melbourne, Sydney, Perth or other cities who attend afestival can be further segmented according to more unique consumption systems.

Introduction

What are the principal associations amongtravellers segmented by demographic vari-ables, their origins, overnight trip destina-tions, trip activities and mode of travel? Cananalysing the possible multiple associationsamong these core trip variables identifyunique tourism consumption systems? Thischapter supports the view that uniquetourism consumption systems (TCSs) can beidentified among overnight leisure trips.

The chapter begins with a small set of core

assumptions and propositions concerningleisure travel behaviour. Second, the methodused to examine the propositions of the studyis described. Third, the findings and limitationsof the study are presented. Finally, implicationsfor additional research and implications fortourism marketing strategy are discussed.

Core Assumptions and Propositions

Following Senge’s (1990) proposition thatunderstanding phenomena increases by stat-



ing core assumptions explicitly, this sectionreviews the core assumptions and proposi-tions driving the study.

Domestic leisure travel is not a given

Domestic tourism constitutes 70% of tourismexpenditures in Australia; it is critical to thecommercial sustainability of the industry.Because domestic leisure travel is discre-tionary, the industry competes with otheruses for consumers’ time and leisure pur-suits, most notably, staying at home to watchTV, gardening, DIY projects, local shoppingtrips, as well as overseas trips. Consequently,dramatic annual shifts in domesticovernight travel behaviour can and dooccur. Throughout the 1990s, domestictourism expenditures included bothdeclines and growth.

During 1998, 74 million overnight tripswere estimated to have been taken inAustralia by Australian residents aged 15years and over; 74% of these visitors hadovernight stays in their own State orTerritory of residence, and 28% of visitorshad an interstate stay. The average overnighttrip length was 4 nights; 81% of the tripswere for leisure and visiting friends and rela-tives (VFR); 19% was business related.Expenditures by overnight visitors amountedto $33 billion in 1998. See BTR (1999) forfurther details.

Tourism marketing often impacts tourismbehaviour

Combining the marketing budgets of spe-cialized government agencies (e.g. TourismNew South Wales, TNSW) and individualenterprises, more than $300 million is spentannually in marketing programmesdesigned to inform and attract Australians,and international visitors, to make leisuretrips in Australia. The core assumptionbeing made here is that tourism-marketingprogrammes are effective in generatingleisure travel behaviour that would not haveoccurred otherwise.

Tourism marketing works at two levels

The explicit primary objective for each imple-mented tourism marketing strategy is to influ-ence consumers to visit/buy a specificdestination, attraction or event-experience.The implicit secondary objective is toincrease leisure travel as a discretionary activ-ity available in the minds of consumers. Thus,the core assumptions include identifying twotypes of tourism marketing influences: abrand-level (e.g. specific destination) influ-ence and a product category (i.e. tourismleisure behaviour versus buying other servicesor durable products) selection influence.

Australians vary dramatically in their leisuretime and tourism behaviours

Australian households may be divided use-fully into three primary travel segments: (i)the segment with a strong propensity to nevertravel overnight away from home; (ii) the seg-ment with a high variance in the number ofovernight leisure trips annually; (iii) the seg-ment consistently making one or moreovernight trips annually. Segments (ii) and(iii) represent the primary qualified marketsto consider for designing tourism marketingprogrammes.

Tourism marketing impacts vary by micro-segment

Within segments (ii) and (iii), identifyingmore micro-behavioural segments is useful,such as tourists making self-directed tours ofvineyards, tourists travelling to attend per-forming arts productions and tourists plan-ning their trips around golf outings.Substantial evidence supports the proposi-tion that the impact of marketing strategies isgreater in attracting some micro-segmentscompared to others. For example, marketingprogrammes to influence visits to wine-grow-ing regions may have greater marginalimpact compared to the marginal impactdelivered by marketing programmes to influ-ence overnight golf outings (i.e. overnight

76 E. Cowley et al.


golfing trips is a more developed tourismmicro-industry compared to overnight wine-region trips; consequently, implementingeffective marketing strategies to influencewine-region travel may be more critical forthis industry compared to the overnight golf-ing-related industry).

Also, the tendency may be more wide-spread among vineyard visitors to view suchtrips as a one-time only trip (e.g. ‘We’ve donethe Hunter Valley’) compared to a golfingtravellers’ proneness for repeat visits to thesame destination. Thus, the vineyard tourismindustry must find ways to market variedexperiences (e.g. wine-tasting-related festi-vals) to overcome the assumption by some vis-itors that a one-off trip is sufficient.

Domestic overnight leisure travel includesunique consumption systems

Particular activities sought by travellers tendto be found in certain destinations more sothan other destinations. Certain traveller ori-gins are associated positively with a few spe-cific destinations, negatively with trips toother destinations, and not at all with otherdestinations. For example, travellers fromSydney may be prone to visiting South CoastNSW but not Canberra, while South CoastNSW residents may be prone to visit bothCanberra and Sydney. Fine dining versustakeaway eating is a more frequent activity inassociation with visiting certain destinationscompared to other destinations. Travellerssegmented by demographic characteristicsengage in specific trip consumptions morethan other trip consumptions, for example,households travelling with children under 15years old tend to go to beaches more oftenthan households without young children.

Consequently, certain destinations relateto specific origins, trip activities and demo-graphic profiles to form trip consumption sys-tems (Woodside and MacDonald, 1994).Specific trip consumption systems representthe strongest markets for attracting visitorsfor particular destinations. Thus, destinationmarketing strategists may be likely to increasethe effectiveness of tourism marketing pro-

grammes by focusing on ‘gilding the lily’, i.e.promoting TCSs where the destinations areknown to deliver well to travellers.

Given these assumptions, the objective ofthe study focuses on describing the TCSs thatassociate consumer demographic segments,origins, trip mode, activities and destinations.Figure 7.1 summarizes this focus by showing ageneralized hypothetical TCS linking a givenorigin with one destination for one travellerdemographic segment using one mode oftravel and two specific destination activities.

While the core proposition that uniqueTCSs relate to specific destinations mayappear obvious, the value of the analysis is inverifying and deepening the TCS view foreach of the major origins and destinations inAustralia. For example, what are the origins,destinations, trip modes and other travelactivities that relate to the Australianovernight golf traveller? Having a completedescription of such a TCS, and others, offersboth theoretical and practical advances tounderstanding traveller behaviour.

The propositions shown in Fig. 7.1 includethe following points:

● P1: domestic overnight travellers seg-mented by their origins have uniquedemographic profiles. For example, theproportion of Australian domesticovernight travellers (ADOTs) high-incomehouseholds is substantially greater forSydney and Melbourne residents in com-parison to Alice Springs and Orange resi-dents.

● P2: ADOT’s demographic segments varysystematically by principal activitiesengaged in during their trips. For example,families with young children often go tothe beach; older couples with no childrenliving at home often engage in fine dining.

● P3: Specific modes of travel link particularorigins and destinations among ADOTs.For example, air travel dominates travelfrom Tasmania to NSW among nearly allADOTs.

● P4: Destinations vary systematically by theprincipal activities engaged in by ADOTs.For example, productions of the perform-ing arts usually occur in large metropoli-tan areas; however, exceptions to this view

Domestic Leisure Traveller Consumption Systems 77


do occur (e.g. the 2-week music festivalheld each summer in rural Finland drawssubstantial numbers of Finnish and inter-national visitors).

● P5: Certain principal activity by someADOT segments results in specific sec-ondary activities. For example, fine diningmay be associated positively with attendingthe performing arts.

While each of these propositions mayappear to be intuitive, unique and valuableinsights for tourism theory and managementstrategy come from examining the proposi-tions in particular contexts, for example,comparing TCSs for Perth, Melbourne,Hobart, Sydney and other origins–destina-tions among ADOTs. Also, empirical evidencefinds strong support that humans tendtoward being highly confident about theirviews of what is ‘intuitively obviously true’,even though these views are mistaken (seeGilovich, 1993, for a review of the evidence).In doing so, applying the TCS method offersthe opportunity for cost-effective and tar-geted marketing strategies.

Method

For this study, additional analyses were con-ducted on the unit record data of overnightdomestic travellers (n = 27,653 households)in the Australian National Visitor Survey(NVS) 1998 study on the characteristics andtravel patterns of domestic tourists. In theNVS study, interviews were done continuouslythroughout the year, using computer-assistedtelephone interviewing. ‘The NVS is an ori-gin-based survey. In simple terms, this meansthat respondents were interviewed in theirhomes and details were collected about theirrecent travel’ (BTR, 1999, p. 67).

The present study focuses on overnighttrips, that is, trips involving a stay away fromhome for at least one night to less than 12months duration, at a place at least 40 kmfrom home. Questions about the overnighttrips in the NVS study focused on the mostrecent completed trip taken within the 4weeks prior to the interview.

Findings are generalized to the residentpopulation of Australia in the following esti-mates. For 1998, overnight visitors included

78 E. Cowley et al.

P2a

P4

P5

P2b

P3a

P3bP1

Visitordemographicsegment Q

Principal destination or

tour route

Principal travelactivity

Secondaryactivity

Mode oftravel related toprincipal travel

activityOrigin R

Fig. 7.1. Theory to tourism consumption systems. P1, origins vary by demographic characteristics (e.g.average incomes are higher in larger versus smaller cities); P2, traveller demographic segement vary in theirpreference for different travel destinations and activities; P3, the mode of travel used often varies amongspecific origins and destinations; P4, specific destinations are associated with specific principal visitoractivities; P5, engaging in a principal activity leads to participating in specific secondary activities.


74,000,000 travellers who spent 293,456,000nights away from home (i.e. the number oftravellers and nights away reflect multiple triptaking). Additional details are summarized inFig. 7.2 for overnight trip origins, destina-tions and accommodations.

The top five leisure activities by overnightresident visitors estimated for Australiainclude the following: VFR (44%); go tobeach and watersports (25%); visit nationalparks, bushwalking, rainforest walks (14%);pubs, clubs and discos (12%); and other out-door activities (9%). The shares for eachactivity vary somewhat by state and territory;for example, going to the beach and water-sports dominates Queensland overnight visi-tors (46%), and VFR is second forQueensland (38%). See BTR (1999) for fur-ther details.

Coding procedure

Not provided in the BTR (1999) are somepossible links suggested by Fig. 7.1. For

example, from Fig. 7.1: what origins are asso-ciated with what overnight trip activities?Thus, are the main reasons associated withMelbourne originating visitors going toSydney, as well as other origin–reason–desti-nation links, not provided in BTR (1999)?

To learn the principal associations amongall the relationships shown in Fig. 7.1, dummycodes (0, 1) were created for several originsand destinations in each State and Territory,as well as type of accommodations, trip activi-ties and demographic levels (e.g. having chil-dren under 15 years old on the trip wascoded 1 versus not having children under 15years old on the trip coded 0). Householdincome in the top 20% of all Australia house-holds was coded 1 and lower incomes werecoded zero.

Quick clustering data analysis

Using the dummy-coded data for origins,destinations, activities, accommodations andmode of travel, the quick clustering proce-


0.03

State/Territory origins State/Territory destinations Accommodations used

NSW

VIC

QLD

SA

WA

TAS

NT

ACT

NSW

VIC

QLD

SA

WA

TAS

NT

ACT

0.01

0.21a

0.12

0.120.04

0.05

0.04

0.03

0.05

0.04

0.07

0.01

0.005

0.01 0.01

0.01

0.01

0.04

0.020.03

0.010.09 0.02

0.06

0.08

0.02

0.01

0.140.03

0.04

0.08bHotel, resort, motel,motor inn

Guest house, B&B

Self-catering cottage,apartment

Caravan park, commercialcamping ground

Caravan or campingon private property

Friend’s or relative’sproperty

0.01

Fig. 7.2. Main findings of the Australian 1998 NVS study: estimated shares of total visitor nights (n =293,456,000). aRead 21% of the total 293,456,000 visitor nights were stays in NSW from trips originating inNSW. bRead 8% of the total 293,456,000 visitor nights were stays in hotels, resorts and motor inns locatedin NSW. Largest percentages are shown in bold. Source: various tables in BTR (1999).


dure recommended by Kamen (1970) wasused to learn the main effects among therelationships shown in Fig. 7.1. This proce-dure includes creating maps of the principal(statistically significant) positive and nega-tive associations found in the correlationmatrix among the variables understudy.Using this method, information is learnedabout possible three-way and four-way associ-ations among the core variables shown inFig. 7.1.

A technical note: use of correlation analysis

Correlation analysis is notorious in behav-ioural research for underestimating effectsizes of relationships among variables(McClelland, 1998). A very few statistical out-liers can reduce a highly significant correla-tion from 0.30 to 0.05, whereby both r-valuesare significant statistically at P less than 0.000.A correlation of 0.05 may be often consideredincorrectly as not noteworthy (Gladwell,1996, 2000). However, if the correlationanalysis resulting in an r estimate of 0.05 withP <0.000 is redone using group level data(Bass et al., 1968), the resulting group-level-based r value is often 0.40 or higher. The keypoint of this discussion is that seemingly lowbut highly significant r values are, in fact,highly significant and of substantial practicalimportance.

The view is often expressed that low, buthighly statistically significant, correlations, aredue mainly to the use of large sample sizes.This view is inaccurate and misleading(McClelland, 1998). The more useful mentalmodel to adopt is that all highly statisticallysignificant correlations demonstrate relation-ships worth further exploration.

These points are readily confirmed fromthe use of cross-tabulation analysis in compar-ison with correlation analysis. However,because the focus of this study is on exploringthe mosaic of relationships among sets of vari-ables, correlation statistical results are pre-sented. The reader is encouraged to explorethe significant relationships discussed ratherthan to apply the mistaken assumption thatlow correlations indicate rather unimportanteffect sizes.

Findings

All five propositions discussed and summa-rized in Fig. 7.1 received strong empiricalsupport. Evidence in support of the fivepropositions is discussed below as well assummarized in the remaining figures in thisreport.

Associations with principal trip activitiesand main destinations visited for ADOTsorganize the findings. The findings start withthe most frequently reported leisure activity(39%) among ADOTs whose main trip pur-pose was holiday/leisure travel. The esti-mated total of 33,092,000 holiday/leisureovernight visitors represents 45% of allovernight travellers by Australian domesticresidents in 1998. Going to the beach is themost often reported activity for overnight hol-iday/leisure travellers. (For all domesticovernight trips, going to the beach is the secondmost reported activity (25% of all trips); thetop activity is visiting friends or relatives(VFR, 44% of all overnight trips).)

Reading the attached figures

Figures 7.2–7.11 provide a visual tool showinglinkages among origins, destinations, travellercharacteristics and selected destination activi-ties. All correlations shown on the linkages(i.e. arrows) are significant statistically.

Main trip activities

Beach

Figure 7.3 presents a summary of significantfindings for going to the beach as a main rea-son for ADOTs. As expected, travel to thebeach is most common in the summermonths (r = 0.07, P <0.000). ADOTs travellingto the beach are most likely to originate theirjourney from Melbourne and Brisbane. Infact, Melbourne is the only city destinationdirectly associated with the beach as anovernight trip activity. Predictably, visitors tothe beach typically have children under 15 (r= 0.05, P <0.000). The most popular beachdestinations include the following:

80 E. Cowley et al.


● the Gold Coast (r = 0.12)● the Sunshine Coast (r = 0.11)● and the south coast of NSW (r = 0.06).

Visitors do not report that they go to largercentres to visit the beach. Visitors to the GoldCoast and the Sunshine Coast were fromDarling Downs and Brisbane in particular.Visitors to the Gold Coast were the onlyADOTs to report significant use of travel pack-ages. They are also most likely to stay in paidaccommodation, and report that they partici-pated in ‘no other activities’ during their stay.The question is whether they choose not toengage in other activities, or could not find,or did not seek out, other activities of interest.Visitors to the Gold Coast may have travelledelsewhere to engage in other activities, forinstance, ADOTs from Gold Coast travellingto the Sunshine Coast, perhaps with touringor fishing in mind. Visitors to the south coastof NSW were from Sydney and Canberra. SeeFig. 7.3 for further details.

Combining a trip to the beach with fishingis a separate type of trip. Figure 7.3 also showsdetailed findings for beach/fishing-relatedovernight trips. Beach/fishing trips are takenin the south coast of NSW and the SunshineCoast, as well as the south west of WA.

The most likely origin for travellers to thesouth west of WA is Perth. The south westcoast of WA was the only beach destinationassociated with other activities unrelated tothe beach such as visiting parks or wineries.Visitors are also likely to travel during thesummer, and are unlikely to incorporate visit-ing friends into the same trip. Visitors lookingfor a combination of beach and fishing areunlikely to look for these activities in Sydneyor Melbourne.

Strategic implications. A key driver for coastaldestinations outside of the urban centres is avisit to the beach. Beach destinations such asthe south coast of NSW draw successfully from


Fig. 7.3. Activity – beach (correlations ≥ |0.05|; for all correlations: P <0.01). �- - - -� Negative association; �—� positive association.


the Sydney and Canberra market. On the otherhand, large urban centres are less successful atdrawing on ADOTs to visit their cities with thebeach as an objective for the trip.

The Gold Coast is associated with packagetrips, while visitors to the Sunshine Coast andthe south coast of NSW are less likely to orga-nize package tours or pay for accommodationon their trip. This may provide an opportu-nity for the Sunshine Coast and the southcoast of NSW to promote package holidays,particularly since ADOTs do not identify visit-ing friends as an objective for the trip.

Wineries

Figure 7.4 summarizes the major associationsrelated directly and indirectly to visitingwineries as an overnight trip activity. Visitorstravelled to the Hunter Valley NSW and southwest WA to visit the wineries. These trips weremade from local destinations in particular;Sydney-siders travelled to the Hunter Valleyand ADOTs from Perth travelled to southwest WA. Neither destination was associatedwith air travel. Visiting wineries is not cur-rently a strong enough driver for domestic

travellers living outside of driving distancefrom the wine-growing area. Counter to intu-ition, the Barossa Valley and other wine-grow-ing areas around Adelaide were not identifiedas a significant overnight destination, norwere areas in Victoria.

In the Hunter Valley, visitors only went tothe wineries, while in south west WA, visitorsreported going to the beach, fishing and visit-ing parks as well as visiting wineries. TheHunter Valley successfully draws ADOTs fromSydney. It is recommended to continue allo-cating promotional resources to the Sydneymarket. Markets requiring air travel, such asBrisbane or Hobart, may not be reasonablyconsidered as potential for the winery dis-tricts. ADOTs from Canberra are currentlynot visiting the Hunter Valley although theyoften visit Sydney with other travel objectives.This may be an untapped market.

Strategic implications. Wineries do notappear to be considered an overnight destina-tion on their own. Other activities at the sameor nearby locations should be promoted aspart of a trip to the wine-growing areas, toencourage travellers to stay overnight.

82 E. Cowley et al.

Fig. 7.4. Activities – wineries (correlations ≥ |0.05|, P <0.01). �- - - -� Negative association; �—� positiveassociation.


Areas such as the Barossa Valley and ClareValley have not been successful at convincingthe ADOTs, particularly those living inAdelaide, to spend at least one night at thesedestinations. It is recommended that thewineries be used as the key driver for thesedestinations supported by other ‘non-winery’activities, such as dining out.

Dining out

Figure 7.5 shows detailed results for diningout as an overnight trip activity. ADOTs travel-ling to large urban centres such as Sydneyand Melbourne participate in dining out.Contrary to Melbourne’s reputation as theculinary centre of Australia, Sydney is themost strongly associated with this activity.Visitors to Sydney are from Melbourne,Canberra and the south coast of NSW. Thegreatest draw is from Melbourne andCanberra. People originating in Melbourneare most likely to travel with the objective ofdining out in mind. They are more likely tobe in a higher-income bracket, and are more

likely to travel by air. ADOTs to Melbourneare from Gippsland, Adelaide and Tasmania.Visitors travelling from Tasmania are alsolikely to travel by air.

Dining out is negatively correlated to visit-ing friends (r = �0.12, P <0.000). Dining out isalso negatively correlated with beach/fishingdestinations such as the south coast of NSWand the Sunshine Coast. An urban setting is acritical determinant for the dining out activityfor ADOTs. Sydney is very successful at attract-ing ADOTs from nearby Canberra and thesouth coast of NSW, as well as Melbourne. It isrecommended that they continue allocatingpromotional resources to these centres.

Melbourne is also successful at attractingADOTs from Adelaide and Tasmania. Again,promotional resources should continue to bespent encouraging these markets.

Brisbane is not associated with dining out.Since there is a number of beach destinationsnear Brisbane, there may be potential in pro-moting dining out by promoting somethingassociated positively with dining, such asgoing to the beach.


Fig. 7.5. Activity – dine out (correlations ≥ |0.04|, P <0.01). �- - - -� Negative association; �—� positiveassociation.


Strategic implications. There is a definitelack of dining out activity in beach locationsand winery destinations. The possibility forstimulating the start-up of such relationshipsamong ADOTs – combining winery tours anddining out or beach and dining out – may beworthwhile for field testing.

Visiting friends and relatives

Figure 7.6 summarizes key relationships withVFR as the main purpose of domesticovernight trips. ADOTs travel to large urbancentres when visiting friends or relatives. Themost common destinations are Melbourne,Perth and Brisbane. The trips are unlikely tooccur during the school holidays in the firstquarter of the year (r = �0.07, P <0.001).Those visiting friends are unlikely to havechildren under 15 and are not in a high-income bracket. It is unlikely that they would

travel by air to the destination, unless they areoriginating in Tasmania.

Travellers visiting friends in Melbourneoriginate from Adelaide, Tasmania andGippsland. Although Melbourne is a commondestination for dining out, this activity is nega-tively correlated with visiting. Travellers toPerth usually are likely to be travelling locally.There is also a segment that travels locallycombining visiting friends with a shoppingtrip. Perth was the only destination where trav-ellers identified shopping as part of the objec-tive of the trip. Finally, those travelling toBrisbane originate in the Gold Coast andDarling Downs.

Strategic implications. The associations sum-marized in Fig. 7.6 may confirm the marketingstrategy that expending marketing resourceson promoting VFR domestic overnight travelmay have lower performance payoffs com-

84 E. Cowley et al.

Fig. 7.6. Activity – visit friends (correlations ≥ |0.05|, P <0.01). �- - - -� Negative association; �—� positiveassociation.


pared to promoting other activity-relatedovernight trips – such as going to the beach.VFR-related overnight travel is indeed relatedto avoiding spending money for other trip-related activities, such as dining out and airtravel. Thus, even though some tourism mar-keting destinations are focused on stimulatingVFR trips (e.g. Pennsylvania’s ‘You have afriend in Pennsylvania’ campaign), such a posi-tioning strategy is not recommended for pro-moting Australia overnight domestic travel.

Destinations

Melbourne

Figure 7.7 summarizes the results for vari-ables for Melbourne-related trips. ADOTs arelikely to travel to Melbourne from Gippsland,Adelaide and Tasmania (Hobart, Launceston).It is unlikely that ADOTs often originate fromgreater distances (i.e. Perth or Brisbane).

ADOTs travelling to Melbourne fall intotwo distinct segments:

● those travelling for a culinary experience,and

● those travelling to visit friends.

The dining out segment is of a higherincome and is more likely to use air travel.The visiting segment is less likely to have ahigh income and is less likely to travel by air.The visiting segment is also unlikely to havechildren under 15.

Visits to Melbourne are not associated sig-nificantly with the objectives of going to thebeach, going to parks, fishing or playing golf.

Strategic implications. Promotional resourcessupporting the restaurant experience inMelbourne has been effective and should con-tinue to be supported. Any attempt to pro-mote an outdoor experience has not been assuccessful and may not be a viable strategy.


MELBOURNE

Children under 15 Air travel

Melbourne

Perth

Adelaide

Brisbane

Gippsland

Hobart

Launceston

Darling Downs

ORIGINS

Dine out

Visit friends

Beach/fish

Beach

Parks

Golf

–0.07

–0.05

0.10

High income0.07

0.07

Disability–0.07

–0.07

–0.04

–0.05

0.11

0.13

0.17

–0.07

–0.10

–0.12

–0.04

–0.04

–0.11

–0.07

0.04

0.05

0.04

–0.06

–0.07

0.05

–0.05

Destination – Melbourne(correlations ≥ 0.04

Negative association

Positive association

Fig. 7.7. Destination – Sydney (correlations ≥ r = |0.04|, P <0.01). �- - - -� Negative association; �—�

positive association.


Perth

Figure 7.8 summarizes the results for Perth.As Perth is one of the most isolated cities inthe world, it is not surprising that ADOTsoriginating from the other side of the coun-try are uncommon. The most likely ADOTorigin is Perth itself. The activities com-monly reported are shopping and visitingfriends. Those visiting friends only areunlikely to do so in the first quarter of theyear and are not likely to have childrenunder 15.

Although ADOTs from Perth go to southwest WA for recreation, there is no indicationthat ADOTs originating from the south westgo to Perth for dining out or the performingarts. Since Perth is the only large urban cen-tre in this area of the state, it is surprising thatthere are not more overnight trips made forthe amenities of the city.

Adelaide

Figure 7.9 summarizes the results for Perth.ADOTs travelling to Adelaide are likely to origi-nate their journey either locally, from Adelaide,or from Melbourne. They are likely to be visit-ing friends, and are not likely to be travelling toAdelaide in order to use the beach. TheADOTs are unlikely to have children under 15.

Adelaide is set in wine country. It is sur-prising that more ADOTs are not travelling toAdelaide to see the wine areas in the Barossaand Clare Valley, amongst other wine areas.The strategic implication in that there ispotential here to increase the number ofovernight stays by promoting wine tours.

Brisbane

Please refer to Fig. 7.10 for the results forBrisbane. ADOTs come from Gold Coast and

86 E. Cowley et al.

Fig. 7.8. Destination – Perth (correlations ≥ |0.04|, P < 0.01). �- - - -� Negative association; �—� positiveassociation.



Fig. 7.9. Destination – Adelaide (correlations ≥ |0.04|, P <0.001). �- - - -� Negative association; �—� positiveassociation.

Fig. 7.10. Destination – Brisbane (correlations ≥ |0.04|, P <0.01). �- - - -� Negative association; �—� positiveassociation.


Darling Downs to visit friends and shop inBrisbane. Although air travel is associatedwith visits to Brisbane, it is not likely forovernight trips when the objective of the tripis to visit friends.

Brisbane is the largest centre along thesouthern Queensland, northern New SouthWales coast. It is recommended that Brisbanepromote the amenities of urban life includingrestaurants, entertainment and shopping.There are many tourists visiting the GoldCoast and the Sunshine Coast that may beinterested in dining out and shopping.Currently these centres have not satisfiedthese needs amongst visitors.

Sydney

Figure 7.11 shows the results for Sydney.ADOTs travelling to Sydney tend to originatefrom Melbourne, Canberra and the southcoast of NSW. Generally, they are of a higherincome bracket. They are interested in dining

out and the performing arts. They do nottravel to Sydney for the beaches, the fishingor golf.

The key drivers for Sydney destinationADOTs are the urban lifestyle, culinary expe-riences and the performing arts. It is recom-mended that promotional resources continueto be allocated toward these types of activity,as they are most strongly associated withSydney. It is not recommended to emphasizeonly the city beaches to the ADOTs as theyare likely to look for beach type vacations out-side of the larger urban centres.

Limitations

A complete test of the core TCS propositionwould include segmented travellers intounique TCSs. This report has focused on con-firming that statistically significant multiplepath associations do occur as set forth in theTCS framework. Additional empirical work is

88 E. Cowley et al.

Fig. 7.11. Destination – Sydney (correlations ≥ r = |0.04|, P <0.01). �- - - -� Negative association; �—�

positive association.


necessary to examine the sizes and types ofTCSs related to specific origins, destinations,travel modes and trip activities. Suggestionsfor such research work are described in thenext section of this report.

The findings and discussion are illustrativeand exploratory, rather than a completeanalysis of the potential for data analysis ofTCSs. Within the Australian National VisitorSurvey database used for this report, substan-tial scope remains for deeper descriptions ofTCSs related to specific destinations, activitiesand trip purposes.

Implications for Further Research

A core proposition for doing the researchwork reported here is that understandingdomestic overnight travel behaviour requiresexamining TCSs rather than only focusing ontwo or three variable relationships. The theo-retical logic of focusing on tourism trips as aunit of analysis rather than variable relation-ships can be applied through field studies ofTCSs. Of course, the more variables includedin the operational definition of TCSs, thegreater the complexity in identifying TCSmicro-segments.

For example, assuming ten origins, fivemain trip purposes, 20 trip activities, and fivemain modes of travel and ten destinations,results in a total of 50,000 possible combina-tions (i.e. potential TCSs). However, a largedata set (e.g. cases >20,000) may be used toconfirm the assumption that far less than themaximum possible TCSs actually occur. Also,about 20–40 TCSs are likely to represent morethan 50% of the total overnight visitors for anyone destination (e.g. NSW as a destination).

The comparative method is a caseresearch method that can be applied to largedata sets (Ragin, 1989). The comparativemethod may be used to identify ‘pure’ TCScustomer segments. This approach, as devel-oped by Ragin, applies Boolean algebra toidentify the multiple paths that lead from oneevent to a given outcome. For example, whatare all the multiple paths leading fromMelbourne as an origin to Sydney as a desti-nation that includes a mode of travel, pur-pose of trip and leisure activities? Applying

the comparative method to a many groupsolution (e.g. g = 100, where g is the numberof groups the researcher requests in the solu-tion) in cluster analysis with binary codeddata, results in a number of pure-type groups.A pure-type group is a series of ‘yes’ responses(coded as 1s) that link a given origin to agiven destination via a given purpose–activ-ity–travel mode path. Such a data analysisapproach is anathema to achieving the goalof a few (g <7) group solutions that fit thedominant logic in positivistic research.

Consequently, the suggested use of thecomparative method represents a heretic viewof applying a case approach to large, quanti-tative databases. This suggested view complieswith the recommendation by the executivesat www.thinktools.com to ‘embrace complex-ity’ (e.g. achieving wisdom about what drivesrelationships in specific contexts, rather thanrestricting data analysis to achieve a limitednumber of parsimonious models of predic-tion and explanation).

This suggested research approach departsfrom the dominant logic (i.e. explaining vari-ance using analysis of variance proceduressuch as F-tests and regression analysis with anemphasis on achieving parsimony). The alter-native logic of the comparative method is touncover the complexity and manage ambigu-ity in the data rather than deciding on a lim-ited number of solutions (e.g. two to threeregression models).

Thus, for tourism behaviour research,applying the comparative method in studiesof tourism behaviour is likely to result inuncovering 100+ major origin–purpose–mode–activity–destination paths (e.g. 5–20per destination) worthy of empirical discus-sion. Such a case-oriented research method isanalogous to testing mostly for five-way inter-action effects rather than testing for maineffects via ANOVA (i.e. analysis of variance).

Implications for Tourism MarketingStrategies

Development and empirical testing of theo-ries of TCSs is a useful approach for the mar-keting strategist who wants to examinespecific paths (or linkages) between customer



origins, destinations, activities, trip purposesand activities. The identification of TCSs rep-resents a useful micro-segmentation viewleading to designing tourism destination–attractions–experiences sought uniquely byeach micro-segment.

The widespread adoption of Internetmarketing strategies is permitting rapid indi-vidualizing of product designs and promo-tional messages to meet the preferences ofprospective customers. Thus, the study ofTCSs can result directly in powerful market-ing strategies when customers go online andidentify what they are looking for, e.g. in adestination. Here is an application of this

marketing strategy: in 2001, Amazon.comprovides each customer with an individual-ized product-offering link after the customercompletes a very limited search at theAmazon.com website.

Thus, a tourism marketer might suggest afew complete TCSs after a prospective cus-tomer completes a limited number of clicks atthe marketer’s website. Given the array ofmultiple trip-related decisions that travellersmust complete for a TCS, the tourism mar-keters which perform best at applying thisstrategy to enable customers to ‘design-your-own TCS’ are most likely to be successful inconverting ‘lookers’ into buyers.

90 E. Cowley et al.

References

Bass, F.M., Tigert, D.J. and Londsdale, R.T. (1968) Market segmentation: group versus individual behav-iour. Journal of Marketing Research 5, 268–272.

BTR (1999) Annual Results of the National Visitor Survey. Bureau of Tourism Research, Canberra.Gilovich, T. (1993) How We Know What Isn’t So: The Fallibility of Human Reason in Everyday Life. Free Press,

New York.Gladwell, M. (1996) The tipping point. The New Yorker 72, 32–39.Gladwell, M. (2000) The Tipping Point: How Little Things Make a Big Difference. Little, Brown & Co., Boston.Kamen, J. (1970) Quick clustering. Journal of Marketing Research 7, 199–204.McClelland, D.C. (1998) Identifying competencies with behavioural–event interviews. Psychological Science

9, 331–339.Ragin, C. (1989) The Comparative Method. University of California Press, Berkeley.Senge, P. (1990) The Fifth Discipline. Doubleday, New York.Woodside, A.G. and MacDonald, R. (1994) A general systems framework of customer choice processes for

tourism services. In: Gasser, R. and Weiermair, K. (eds) Spoilt for Choice. Kultur Verlag, Vienna, pp.30–59.

Woodside, A.G. and Dubelaar, C. (2001) A general theory of tourism consumption systems. Journal ofTravel Research 40, 120–132.


Chapter eightTourist Activity Planning in Congested Urban TourismEnvironments: Towards a Game-theoretic Model andDecision Support System

Qi Han,1 Benedict G.C. Dellaert,2 W. Fred van Raaij3 and Harry J.P. Timmermans1

1Department of Urban Planning, Faculty of Architecture, Building and Planning, EindhovenUniversity of Technology, Eindhoven, The Netherlands; 2Department of Marketing, Faculty ofEconomics and Business Administration, Maastricht University, Maastricht, The Netherlands;3Department of Economic and Social Psychology, Faculty of Social and Behavioural Sciences,Tilburg University, Tilburg, The Netherlands

Abstract

Urban tourism has positive effects on the city such as generating financial support and improving the city’satmosphere, but may also have negative impacts such as overuse of resources. In this chapter, an activity-based approach to tourists’ behaviour analysis is combined with a game-theoretic methodology to study theinteraction between individual tourists, as well as the interaction between tourists and the tourism informa-tion offices. We postulate that, when integrated in an Internet-based information system, this approachcan support urban authorities to achieve sustainable tourism development, and also support individualtourists planning their urban day trips as a customized experience.

Introduction

Understanding needs, preferences and deci-sion making processes of consumers is a keyissue of successful planning and marketing. Inthe domain of urban tourism, it is especiallyrelevant to know urban tourists’ preferencesfor different possible combinations of alterna-tives, because previous research has shown thaturban tourists typically like to combine severalactivities when they make city day trips(Dellaert et al., 1995). Moreover, urban

tourism behaviour often takes place in highlycomplex urban environments that incorporatemultiple tourism facilities and services, whichare used by many tourists simultaneously, oftenresulting in congestion. Therefore, it is desir-able to develop models that allow one toinclude combinations of alternative activitiesas well as interdependencies among decisionmakers. Such models can improve the preci-sion of predicted effects in tourism behaviourand thus further optimize the usage of the city.A promising solution might be an activity-



based approach to tourists’ behaviour analysiscombined with a game-theoretic methodologyto study the interactions between individualtourists as well as the interaction betweentourists and the urban tourism information(UTI) offices. This chapter discusses the con-ceptualization of such a model and how itcould be implemented in an experimentalwebsite to test its validity.

This chapter starts with describing theresearch problem context, focusing on themost relevant approaches (Section 2). It con-tinues with presenting the conceptual frame-work in terms of tourism analysis, tourismplanning and economics (Section 3). Then,an activity-based model is defined by specify-ing its support space, data requirements andthe process (Section 4). Subsequently, agame-theoretic model consisting of two dif-ferent games played at aggregate and disag-gregate level respectively is discussed (Section5). Finally, conclusions and discussion com-plete the chapter.

Background

Traditional tourism planning methods inurban development are primarily supply-oriented and focus on constraints and thephysical–spatial structure of the urban envi-ronment. Over the last 5–10 years, manycities have realized that this physical–spatialapproach is insufficient to achieve most pol-icy objectives, and have tended to adoptmore demand-oriented marketing planningapproaches, which have the advantage ofbeing less centralized and more sustainable.This tendency is reflected in the develop-ment and implementation of the city-market-ing approach.

Urban tourism constitutes one importantaspect of city-marketing. Many cities haveconsidered tourists’ visits and expenditures asan important source of income to financeurban facilities and as a key to urban develop-ment and revitalization. Planners in thesecities hope that tourists will not only spendmoney in their city and enjoy urban facilitiesand services, but also bring cross-fertilizationthat will improve other sectors of the urbaneconomy and improve the atmosphere in the

city. Nevertheless, tourists potentially alsohave a negative impact on the city, especiallywhen viewed in the light of urban sustainabledevelopment (Ryan, 2002). Hence, the chal-lenge is how to influence the activities oftourists in such a way that the balancebetween advantages and disadvantages is opti-mized. Unfortunately, UTI offices often face alack of insight into tourists’ responses totourism information, the influence of theseresponses on the functioning of urbantourism and the potential synergism betweenurban facilities. This knowledge is vital forurban authorities to achieve their objectivesin an integrative urban tourism framework(Pearce, 2001).

Previous studies of urban tourists’ choicebehaviour show that urban tourists typicallycombine several activities in one trip. Thisfurther complicates the understanding oftourist behaviour. This makes the so-calledactivity-based approach a potentially veryvaluable one. The activity-based approachmodels individual behaviour in terms ofwhich activity is conducted where, when, withwhom and with which transport mode, in anattempt to replicate the way that individualsarrived at a comprehensive activity-pattern intime and space. This approach has its originsin geography and urban planning, but is nowa very active line in transportation researchand time use research (Timmermans et al.,2002). The advantage of this approach is thatit incorporates all choice dimensions such asactivity choice, location, duration andsequencing, subject to a wide set of con-straints. All these dimensions together reflectthe scheduling of activities in time–space andrepresent a holistic decision process in a com-plex environment. Some advanced activity-based models of activity–travel behaviour arederived from theories of choice heuristicsthat consumers apply when making decisionsin daily life (e.g. Arentze and Timmermans,2000); others are based on the principle ofutility maximization (e.g. Bowman and Ben-Akiva, 1999).

However, to date, activity-based modelshave hardly paid attention to the modellingof interactions between decision makers. Incongested urban tourism environments, it isimportant to understand the interactions

92 Q. Han et al.


between different decision makers in order tominimize congestion and thus optimize theusage of the tourism attractions and facilitiesin the city. Non-cooperative game theorymight be helpful in this respect. This theory,which emanated from studies of games suchas chess or poker, assumes that players haveto think ahead and devise a strategy based onexpected countermoves from the otherplayer(s). Such strategic interactions charac-terize many economic situations, and this the-ory has therefore proved to be very useful invarious types of economic analyses. Theadvantage of this approach is that it is basedon some primitive assumptions about payoffsand strategies, and uses the concept of equi-librium as the principal guideline to predictthe outcome of strategic interactions(Rasmusen, 1990; Scott Bierman andFernandez, 1998). Thus, the interactionsbetween tourists themselves and betweentourists and the UTI office could be studiedby formulating them as special cases of non-cooperative games. Therefore, we decided tocombine the activity-based approach with agame-theoretic method in our model, aimingto investigate tourists’ behaviour and improvethe optimal use of urban facilities.

In practice, the proposed model will needto be implemented in a tourist decision sup-port system to test its usefulness. Internettechnology may be a promising tool in thisregard, because an increasing number oftourists use the Internet to make travel plans.Moreover, Internet technology can be usednot only for marketing purposes, but also for

optimizing general welfare, provided that itis complemented by an appropriate controlmodel. We argue that when integrated in anInternet-based information system, the pro-posed model can support urban authoritiesto achieve sustainable tourism development,and also support individual tourists in plan-ning their urban day trips as a customizedexperience.

Conceptual Framework

Central to our conceptual framework is theunderstanding of interaction mechanisms.Our Internet decision support system isintended to be an interactive communicationplatform for tourists to build themselves a tai-lored day trip programme, and communicatewith the UTI office about their interests andopportunities, based on the aggregate effects.

A learning process, represented by dashedlines in Fig. 8.1, is permeated through thewhole decision support procedure. Thisprocess assists the UTI office in gaininginsight into tourists’ responses to informa-tion, the influence of these responses on thefunctioning of urban tourism in the city, andthe potential synergism between urban facili-ties. Learning in the system is based on obser-vations of the tourists’ input in buildingtailored day trip schedules, tourists’ feedbackon the effect of information when choosingan optimal strategy and tourists’ day trip port-folio choices, which can be used as a basis toperform simulation and evaluation.

Tourist Activity Planning in Urban Tourism 93

UTI office Tourists

Internet web

Urbanistassistant

Touristassistant

Interact with

Interactwith

Interact with

Learningby touristfeedback

Learning bysimulation

Fig. 8.1. Conceptual framework.


As shown in Fig. 8.2, the proposed deci-sion support system will consist of two parts:the tourist assistant and the urbanist assis-tant. The system includes an experimentalwebsite where tourists are requested tochoose between a set of activity alternativesthat the website suggested, and to decidewhether they are satisfied with the schedulethat the website offered. The acceptedschedules together with tourists’ personalinformation will be recorded by the touristassistant. The urbanist assistant will use thesedata to conduct a simulation analysis, so asto decide the optimal strategy for eachtourist and for the total urban tourism net-work. Thereafter, this strategy will generateinformation presented on the website toeach tourist.

These lead to two loops in the proposedInternet decision support system. One isinside the urbanist assistant subsystem and isimplemented in ‘Simulation analysis’ (seeSection 5 for more details), which offers sys-tem decision support, referred to as the‘inner-loop’. The other is interacting with thetourist assistant subsystem, given the informa-tion resulting from the ‘inner-loop’ and theinput of data collection, referred to as the‘outer-loop’. Such an interactive mechanismis useful for the urbanist to predict the ser-vice effectiveness of the urban environmentand for tourists to plan their day-trips morequickly and effectively.

An Activity-based Model – the TouristAssistant

The tourist assistant supports the tourists dur-ing his/her day-trip planning decision mak-ing process. It is developed from anactivity-based model. In Fig. 8.3, the greycoloured text boxes show the tourist’s actions,while the other boxes indicate the touristassistant’s action. After a tourist inputshis/her personal information, the touristassistant gives some activity options with pre-scriptive information at the activity level. Nextthe tourist makes a selection and giveshis/her preferences. Based on theseresponses, the tourist assistant generates aschedule and presents it to the tourist withanticipatory descriptive information at theschedule level. If the tourist is satisfied withthe schedule, the tourist assistant records thedata; otherwise, the tourist assistant returnsto update the activity or schedule suggestionbased on the more precise tourist input. Theactual process of scheduling activities is con-ceptualized as a process in which a touristattempts to achieve a day trip programmethat has the best combination of all travelobjects he/she requires, given a variety ofconstraints that limits the number of feasibleday trip plans. This day trip programme com-prises an ordered sequence of activities andrelated travels between particular locations,with particular start times and durations, and

94 Q. Han et al.

UTI officeActivities, facilities, travelAccessibility distributionStrategy adjustmentEffectiveness prediction

TouristsUser information(character, budget,preference and choice)Day-trip plan

Internet web

Urbanist assistantSystem decision support

Tourist assistantPersonal decision support

Influence information

Simulation analysis Data collection

Schedule generation

Activity option

Fig. 8.2. Web system architecture.


with particular transport modes. By varyingthe activity level and schedule level informa-tion, tourists’ reactions to different sugges-tions can be investigated.

In order to give tourists better sugges-tions, it is important to distinguish betweenplanned and impulse activities. In mostcases, the decision to visit a city is activatedby some planned activities such as visiting aparticular historical destination or a particu-lar shopping area. It is rational to assumethat day trip structures depend primarily onthese planned activities, which are relativelyrestricted in time and certainly space. Incontrast, the impulse activities could be con-tingent upon many conditions and con-straints when tourists are conducting theirplanned activities. These impulse activitiesare more flexible, which gives our touristassistant more flexibility and space to givesuggestions and observe the responses.Hence, the recommendations in our experi-mental website are preceded by two parts.One is about planned activities and givessuggestions such as better conducting times(start time) together with locations, dura-tions and travel modes. The other generatespossible impulse activities and gives sugges-tions such as the most feasible impulse activi-ties together with locations, start times,durations and travel modes, given the touristpersonal information input.

To generate the suggestions, the systemneeds both environmental information andindividual information. Environmental infor-mation describes the available activities, facili-ties and travel options in the following terms:

● Activities: optional facilities where an activ-ity can be performed, average durationrequired for each activity.

● Facilities: includes opening hours, geo-graphical locations of the facilities(absolute location), centrality of differentfacilities (relative location), the time-dependent crowdedness of the facility(customer density and floor space),entrance fee.

● Travel: travel speeds of different mode(more precise information could includetravel speeds between locations at differenttimes of the day according to traffic condi-tions), route attractiveness, monetary cost(car, public transport and walking).

Individual information will be obtainedfrom tourists’ input, which contains threepieces of information.

● Socio-demographics, psychographics, exper-ience for modelling the segments for whichdifferent suggestions are given.

● Money budget, time budget and plannedactivities together with planned location,for setting the basic constraints.


Information input

Activity options

Option selection

Schedule generation

Schedule confirm

Data record

Age, gender,money and time budget,socio-demographics,psychographics,experiences

Attractiveness description, facilitylocation, average participation duration,opening hours, entrance fee, queuetime, accessibility

Activity, sequence, time window, modechoice, travel party

Activity sequence,time window,generalized cost

Anticipatorydescriptiveinformation(schedule level)

Prescriptiveinformation(activity level)

N

Y

Fig. 8.3. Tourist assistant model.


● Personal preferences for activity types, activ-ity performing facilities and locations, dura-tions and times, and travel modes togetherwith environmental information, for figur-ing out the context-aware suggestions.

At the activity level, a critical processassumption is whether people plan one activ-ity at a time, or schedule several activitiessimultaneously. Following the satisficing prin-ciple (Simon, 1982) and in order to get amore tractable operational model, we assumethat day-trip planning decisions are likely tobe incremental. Only one activity at a time isconsidered or added to the activities list.

We assume that individuals expect toobtain some benefit from each activity in theday. We use the common term utility to referto the net benefits derived from an activity.However, an activity is not taken into accountif the cost of its performance is higher thanthe utility of its performance. Based onLouviere and Timmermans (1992) andGärling et al. (1998)’s research on hierarchi-cal information integration, it will often bethe case that an activity choice entails severalsubordinate judgements or choices, which aremutually dependent, such as duration, loca-tion, travel time, and so on.

At the highest level, a person weighs theutility of performing an activity against thecost of performing it. At the middle level,he/she makes a judgement of the cost(including time pressure and travel cost)and the utility (including motivation andperceived utility). At the lower level, travelcost accounts for location accessibility, traveltime, route benefit, waiting time and timeflexibility, and perceived utility accounts forthe variation of utilities of the activities overtime–space such as attractiveness and crowd-edness. Our tourist assistant will mimic thiskind of human mental process to get themost interesting suggestion for the specifictourist.

After calculation, the activity with thehighest utility will always be the candidate forbeing recommended first and likewise. Noadjustment of duration is made at this phase,i.e. activity level. It is rational to assume that,when the tourist assistant gives ‘activityoptions’ (see Fig. 8.3), the system only takesthe average duration into account. Because at

the first recommendation phase when thetourist tries to get a list of activities, he/shedoes not exactly know how long each activitywill last. It is assumed that if the time avail-able is less than the specified average dura-tion, the activity will not be considered. Ifmore time is available, either there will besome waiting time or another activity will besuggested next. Given the first round of rec-ommendations, a tourist will decide to acceptthe recommendation or not. Accepted rec-ommendations will be taken into accountwhen the assistant gives the next suggestion.This procedure will be repeated until thetime budget and money budget areexhausted.

Having generated the list of the activities,the tourist assistant will give a schedule levelsuggestion by generating the appropriateschedule. From pedestrian destination androute choice studies, we know that differentpeople apparently use different decisionheuristics (Gärling and Gärling, 1988;Timmermans et al., 1992) such as the global-distance minimizing, local-distance minimizingor total-distance minimizing strategy to choosea route, when they have a list of stops to visit.However, these rules only will not completelysolve the scheduling problem. In the case inwhich a pedestrian knows which stores to visitand has decided on which route to follow, theproblem in which sequence to visit the storesstill remains, i.e. the route can still be followedin two directions. The alternative sequentialpatterns are unrelated to the total travel dis-tance. Specific sequence heuristics used hereare the nearest-destination-oriented, furthest-destination-oriented or intermediate-destina-tion-oriented rules. Because of the similaritybetween pedestrian behaviour and day tripbehaviour, we could suppose these heuristicsare also applicable in the case of planning day-trip. Moreover, the pedestrian studies men-tioned above are based on the assumption thatnon-location attributes are irrelevant; in real-ity, it is clear that destinations differ in terms oftheir attractiveness and the context of the per-son and environmental situation. Therefore,the tourist assistant should integrate theseheuristics with context-aware information togive better activities sequence suggestions, ortiming of the activities.

96 Q. Han et al.


In the second phase, the duration of activi-ties will be slightly adjusted by using a utilityfunction of duration time according to per-sonal preference and constraints. Recently,Timmermans et al. (2001) and Joh et al.(2001) proposed a model of activity resched-uling. They described an asymmetric S-shapeutility function of duration time. This func-tion reflects the assumption that for eachactivity there is a warming-up period, duringwhich marginal utility is small, a core periodwith high marginal utility and a cooling-downperiod, in which marginal utility is small.Without discussing the model in detail, it isnoted here that the proposed utility functionis intuitively plausible, and allows for a flexi-ble specification of activity utility. We adoptthis utility function in our system, havingtaken into account the impacts of all choicefactors other than duration in terms of a daytrip based on context-aware information.

The proposed process attempts to mimichow that context-aware information isretrieved and choice options are constructedwhen people perform this task themselves.This increases the probability that the touristwill be satisfied with the suggestion andaccept it. As mentioned before, these sugges-tions will be implemented at two levels – theactivity level and the schedule level – duringthe procedure of helping tourists generatingday trip plans. The principle of generatingrecommendations will be implemented inboth the ‘inner-loop’ and the ‘outer-loop’.

A Game-theoretic Model – the UrbanistAssistant

The interaction between the simulated indi-vidual tourist and the UTI office in the‘inner-loop’ is an iterative process. To simu-late this process, we need to construct somesub-models and understand the relationshipbetween these models, especially how theywork together to represent the interaction(Fig. 8.4). The tourist behaviour model simu-lates tourists’ actions. It takes the informa-tion provided by tourists and systemsuggestions as input and generates optimalschedule choices according to analysis of the‘outer-loop’ data. This results in a set ofschedules, which are input to the distribu-tion-loading model. The distribution-loadingmodel maps the schedules to the urban facil-ities, given facility time-dependent con-straints. The outputs of the distributionloading are time-dependent facility flows,which are input to the facility performancemodel. Then, the facility performance modelproduces economic evaluations and queuingtimes based on facility capacity. The strategyoptimization model uses these results tochoose the strategy, which applies to the sug-gestion generation model. These suggestionsare provided to the tourist behaviour model.The resulting condition is therefore the equi-librium of the distribution–control interac-tion, like the demand–supply equilibrium in


Control side Distribution side

Suggestion generation

Optimized strategyUsage prediction

Measure of effectiveness

Strategy optimization

Facility performance model(evaluation, queuing time, etc.)

Distribution loading(map schedules to facilities)

Tourist behaviour model(schedule choice)

Equilibrium

Fig. 8.4. Iterative process.


economic analysis. The combined distribu-tion–control problem should be interpretedas the following two games being played.

Game I – the tourist distribution game

The first game is played at the disaggregatelevel – the tourist distribution problem, whichis formulated as an n-player non-cooperativegame among the tourists. It is supposed thatthere are n players in the game and they arenumbered from 1 to n. In this case, each playeris a tourist. Each tourist’s information is imper-fect, certain, symmetric and incomplete, whichmeans that tourists do not know other tourists’behaviour (i.e. where they are and what theyare going to do), but none of the tourists havemore information than the others.

To solve this uncertainty in the game, it issuggested that the uncertainty could be dueto the existence of a small amount of uncer-tainty about the opponent’s payoffs. In gametheory, the so-called Harsanyi transformationtreats players who have different payoffs asbeing of distinct types. It proposes that suchgames be modelled by having ‘Nature’ movefirst and ‘choose’ each player’s type. Playersstart out knowing their own type but not thatof their competitors. Players share a commonbelief as to how ‘Nature’ makes its probabilis-tic choice. The resulting new game structureis called a Bayesian game. Following the gen-eral concept of Bayesian Nash equilibrium, astatic Bayesian game can be described by itsstrategic form, which consists of five lists, allof which are common knowledge. In ourcase, we define the five lists as below:

● First is the list of players, which aretourists.

● Second is the list of moves for each player. Alist of moves, one for each player, is a moveprofile and will be enclosed in curly brack-ets {}. A tourist’s move is the set of all avail-able activity schedules in the city. A schedulemeans a series of activities along with theattributes of location, duration, mode andtime window. Each tourist chooses one setout of his/her feasible sets independently.When a tourist makes a choice, he/she doesnot know which sets are selected by othertourists, as if they move simultaneously.

● Third is a list of possible types for eachplayer. Every player knows his/her owntype but may not know the type of theother players. The other players aredenoted by the index �i. A list of types,one for each player, is a type profile and isenclosed within square brackets []. Atourist often does not know the values thatothers place on the possible outcome. Inother words, a tourist may not know theother tourists’ preferences that lead totheir payoffs, and as a consequence totheir moves. We will use an error term todescribe the uncertainty of the perceivedpayoffs and transfer it to the players’ type.

● The players have a common prior beliefabout the probability of each possible typeprofile. The probability that the playerprofile is [t1,…,tN] is denoted P[t1,…,tN].These probabilities comprise the fourthlist. We will use an initial value to checkthe face validation, and later we can usedata to obtain the exact probability value.

● Every move profile {m1,…,mN} and type pro-file [t1,…,tN] result in a payoff for eachplayer. Player i’s payoff will be denoted ingeneral as Ui(m1,…,mN,t1,…,tN). The payoffsmake up the fifth list. The payoff function isan overall utility for this tourist’s schedule.The game tree (Fig. 8.5) gives the extensiveform of an n-player game for a schedulechoice problem. In this form, the touristschoose their schedules as if they happen insequence, but actually when one touristmakes a decision, he/she does not know theothers’ selected schedules. The dashed linethat connects the nodes at each tourist levelindicates that those nodes are indistinguish-able. Each leaf at the end of the tree repre-sents a possible outcome of the game.

In general, player i’s move may depend onhis/her type. Hence, a strategy for player iconsists of a move as a function of player’stype and will be denoted as Si(ti). A list ofstrategies, one for each player, is a strategyprofile and will be enclosed within curlybrackets. Since the players’ types are ran-domly selected, the outcome of a strategy pro-file will be uncertain. We assume the playersare expected utility maximizers and chooseamong strategies on the basis of conditionalexpected utility (EU). The conditional EU of

98 Q. Han et al.


player i of type ti when the players adopt thestrategy profile S = {S1(t1),…,SN(tN)} is

EUi(S,ti) = ∑Ui(S1(t1),…,Si(ti),…,SN(tN),t�i

t1,…,ti,…,tN)•Pi[t�i ti],

where the summation is over all the possibletype profiles of the other players. The defini-tion of a Bayesian Nash equilibrium is (e.g.Scott Bierman and Fernandez, 1998):

A strategy profile S* = {S *1(t1),…,S *

N(tN)} isa Bayesian Nash equilibrium of a staticBayesian game if and only if for every playeri, every type ti of player i, and every alterna-tive strategy of player i, say Si(ti), EUi(S*,ti)≥ EUi(S *

1(t1),…,Si(ti),…,S *N(tN),ti). That

is, the player’s strategy is a best response tothe strategies of the other players, whateverthe player’s type.

That is, each player’s strategy Si solves

maxUi(S *1,…,Si,…,Sn

*)si�Si

If no single tourist can improve his/her pay-off function by unilaterally changing his/herschedule, then it is essentially a BayesianNash equilibrium, in which tourists maximizetheir own individual utilities.

Game II – the tourism distribution–controlgame

The second game is a game at the aggregatelevel between the UTI office and the collec-tive tourists – the tourism distribution–controlproblem. In this game, there is anotherplayer – the UTI office. The purpose of theUTI office is to increase tourist demand andinfluence the use of urban facilities to reachan optimum level of usage. Adding the UTIoffice to the game is not as simple as extend-ing an n-player game to an n+1-player game,because the strategy space and the payofffunction for this additional player differ fromthe rest of the n players.

In order to investigate the second game,we assume that the tourists’ activity schedulechoice behaviour can be characterized by aBayesian Nash equilibrium. This allows us totreat the collective tourists as one player whodistributes the tourism flows to the urbanfacility network. Then, a two-player non-coop-erative game is formed. The two players haveimperfect, certain, asymmetric and completeinformation. The UTI office can choose thebest control strategy to achieve the highest


(1,1

...,1

)

(1,1

...,2

)

(1,1

...,S

n)

(1,2

...,1

)

(1,2

...,2

)

(1,2

...,S

n)

(1,3

...,1

)

(1,3

...,2

)

(1,3

...,S

n)

(Sn,

Sn,

...,1

)

(Sn,

Sn,

...,2

)

(Sn,

Sn,

...,S

n)

Tourist 1

S2 Sn

S1 SnS2

Schedule 1

S1S2

Sn S1S2

Sn

Tourist 2

Tourist n

S1S2

Sn S1S2

Sn S1 SnS2

S1 SnS2

Fig. 8.5. Tourist distribution game extensive form.


system payoff, while taking the anticipatedtourists’ reactions into account. This meansthat this is a two-stage game. First, the UTIoffice sets the strategy. Second, the touristsobserve the result of the strategy settings andchoose their best series of activity schedulesaccordingly. Since the UTI office can solvethe tourists’ optimization problem, theyshould anticipate the tourists’ best reaction toeach possible control strategy in selecting thebest strategy. In the case of duopoly (two play-ers), such a multi-stage game is known in eco-nomics as a so-called Stackelberg game. Theplayer who moves first is a leader and theplayer who moves next is a follower. The UTIoffice’s payoff is a utility function of the over-all tourism distribution flows in the city, andthe collective tourists’ payoff is a utility func-tion of tourists’ schedules. The upside-downtree structure (Fig. 8.6) represents the exten-sive form of this game. The key characteristicsof this multi-stage game of complete informa-tion are that the moves occur in sequence, allprevious moves are observed before the nextmove is chosen, and that the players’ payofffrom each feasible combination of moves iscommon knowledge.

In general, the equilibrium of this game isnot a Nash equilibrium. Rather, the equilib-rium is determined by backward induction.The UTI office first initiates the move by set-ting the control strategy, g. When the touristsmake their move based on their strategy, h, atthe game’s second stage they will face the fol-lowing problem:

maxUh(g, h)h�H

Assume that, for each g in G, the above opti-mization problem has a unique solution.Denote the unique optimal solution by h*(g).h*(g) is the tourists reaction or best responseto the UTI office’s strategy g. Since the UTIoffice can solve the tourists’ optimizationproblem, they should anticipate the tourists’best reaction to each possible control strategyg. Therefore, the UTI office’s problemamounts to:

maxUg(g,h*(g))g�G

The equilibrium of the Stackelberg game canthus be defined as (e.g. Scott Bierman andFernandez, 1998):

In the two-player game between the UTIoffice and tourists, �= {G,H;Ug,Uh}, the com-bination of strategies (g*, h*) is a Stackelbergequilibrium if strategy combination (g*, h*)solves the following bi-level problem:

maxUg(g,h*(g)) g �GSubject to: h*(g) = maxUh(g,h)

h �H

Conclusions and Discussion

This chapter has put forward the concep-tual framework of an Internet decision sup-port system prototype for optimizing theusage of urban tourism facilities, and sys-tematically described the type of modelsthat are defined within the decision supportsystem. It presented an activity-basedapproach to model tourist behaviour inurban tourism, meanwhile addressing theproblem of modelling the interactions atboth the disaggregate level (between indi-vidual tourists’ decisions) and the aggregatelevel (between tourists flows and planningdecisions), by means of non-cooperativegame theory.

The proposed application will not onlysupport individual tourists to plan theirurban day trips as a completely customizedexperience, but also support UTI offices toachieve sustainable tourism development bystrengthening their influence on tourists’activities. The system may contribute toimprove the effective and efficient usage ofurban facilities by synchronizing the supplyand demand for urban tourism productsand services. Of course, it should be real-ized that such a system is demanding interms of coordination of data andresources, and that its potential value alsodepends on the number of tourists usingthe system.

We hope to report on our experienceswith this project in future publications.

100 Q. Han et al.



UT

I offi

ce

Str

ateg

y 1

g2gm

Tour

ists

UT

Iof

fice

g2g1

gmg2

g1gm

g2g1

gmg2

h1h2

hn

g1gm

g2

g1gm

g2

g1gm

h1h2

hn

g1gm

g2

g1gm

g2

g1gm

h1h2

hn

g2

Fig.

8.6

.To

uris

m c

ontr

ol-d

istr

ibut

ion

gam

e ex

tens

ive

form

.


References

Arentze, T.A. and Timmermans, H.J.P. (2000) ALBATROSS: a Learning-based Transportation OrientedSimulation System. European Institute of Retailing and Services Studies, The Hague.

Bowman, J.L. and Ben-Akiva, M.E. (1999) The day activity schedule approach to travel demand analysis.Paper presented at the 78th Annual Meeting of the Transportation Research Board, Washington, DC.

Dellaert, B.G.C., Borgers, A.W.J. and Timmermans, H.J.P. (1995) A day in the city: using conjoint choiceexperiments to model urban tourists’ choice of activity packages. Tourism Management 16, 347–353.

Gärling, T. and Gärling, E. (1988) Distance minimizing in downtown pedestrian shopping. Environmentand Planning A 20, 547–554.

Gärling, T., Kalen, T., Romanus, J., Selart, M. and Vilhelmson, B. (1998) Computer simulation of house-hold activity scheduling. Environment and Planning A 30, 665–679.

Joh, C.H., Arentze, T.A. and Timmermans, H.J.P. (2001) Understanding activity scheduling and reschedul-ing behavior: theory and numerical illustration. GeoJournal 53, 359–372.

Louviere, J.J. and Timmermans, H.J.P. (1992) Testing the external validity of hierarchical conjoint analysismodels of recreational destination choice. Leisure Sciences 14, 179–194.

Pearce, D.G. (2001) An integrative framework for urban tourism research. Annals of Tourism Research 28,926–946.

Rasmusen, E. (1990) Games and Information: an Introduction to Game Theory. Basil Blackwell, Oxford.Ryan, C. (2002) Equity, management, power sharing and sustainability – issues of the ‘new tourism’.

Tourism Management 23, 17–26.Scott Bierman, H. and Fernandez, L. (1998) Game Theory with Economic Applications. Addison-Wesley,

Reading, Massachusetts.Simon, H.A. (1982) Models of Bounded Rationality, Vols 1 and 2. MIT Press, Cambridge, Massachusetts.Timmermans, H.J.P., Hagen, X. van der and Borgers, A. (1992) Transportation systems, retail environ-

ments and pedestrian trip chaining behavior: modeling issues and applications. TransportationResearch Board 26B, 45–59.

Timmermans, H.J.P., Arentze, T.A. and Joh, C.H. (2001) Modeling the effects of anticipated time pressureon the execution of activity programs. Transportation Research Record 1752, 8–15.

Timmermans, H.J.P., Arentze, T.A. and Joh, C.H. (2002) Analysing space-time behaviour: new approachesto old problems. Progress in Human Geography 26, 175–190.

102 Q. Han et al.


Chapter nineComparing First-time and Repeat Visitors’ Activity Patternsin a Tourism Environment

Astrid D.A.M. Kemperman, Chang-Hyeon Joh and Harry J.P. TimmermansUrban Planning Group, Eindhoven University of Technology, PO Box 513, 5600 MBEindhoven, The Netherlands

Abstract

This chapter compares activity patterns of first-time and repeat visitors in a theme park. First-time andrepeat visitors are classified into those who use information and those who do not. A sequence alignmentmethod and chi-squared automatic interaction detection are used to classify the visitors with regard totheir activity patterns and identify predictor variables. The results indicate that first-time and repeat visitorsdiffer in their activity patterns in the park, specifically with respect to the order of activities chosen. First-time visitors follow a strict route in the park, while repeaters have a more diverse activity pattern.Furthermore, the difference between the two groups is reduced when first-time visitors use informationabout the available activities and the spatial layout of the park.

Introduction

The last decades have witnessed a constantflow of publications in tourism researchabout consumer preferences and choicebehaviour. The vast majority of such studieshave dealt with a single facet of preferenceformation or choice behaviour, such as thechoice of destination (e.g. Louviere andTimmermans, 1990; Crouch and Louviere,2000), or the choice of season (e.g. Uysal etal., 1994; Murphy and Pritchard, 1997;Siderlis and Moore, 1998). Research dealingwith complex behaviour, in the sense of multi-ple, interrelated choices is relatively scarce.Examples of such research could be cross-sec-

tional, interrelated choices, such as the simul-taneous or sequential choice of destination,transport and activities (e.g. Dellaert et al.,1998a; Jeng and Fesenmaier, 1998; Tideswelland Faulkner, 1999; Taplin and McGinley,2000), the analysis of preferences of differentfamily members (e.g. Dellaert et al., 1998b),and the analysis of temporally related choices,such as variety seeking (Kemperman et al.,2000, 2002) and habit formation(Oppermann, 1998). The study of such com-plex tourism choice behaviour is of interest,not only because it may better unveil the truenature of such behaviour, but also because itgenerates new challenges in the developmentof preference and choice models, and the



exploration of relatively unexplored methodsand techniques in tourism research.

As part of a wider research programmeexamining complex behaviour in a variety ofapplication domains, the present chaptercompares the interrelated activity patternchoices of first-time and repeat visitors in atheme park. While repeat visitation has beenrecognized as important, this has not beentranslated into a substantial number of thor-ough studies of repeat visitation patterns (e.g.Gitelson and Crompton, 1984; Reid and Reid,1993; Oppermann, 1996). Gitelson andCrompton (1984) were among the first to rec-ognize the importance of the repeat visitorsmarket and its marketing implications. Otherstudies followed. Results indicate that thereare significant differences with regard to com-position and travel behaviour of first-time andrepeat visitors (Oppermann, 1997). His studyis one of the few studies that focuses on dif-ferences in aspects of actual travel behaviourbetween first-time and repeat visitors to a des-tination. First-time visitors were visiting manymore attractions within the destination areathan repeat visitors. Hence, first-time visitorsseem to be more active.

Other studies investigated the differencesin intention to travel to a destination (e.g.Fakeye and Crompton, 1991) or to a visitorattraction (Darnell and Johnson, 2001) andin information needs. First-time visitors havelimited knowledge about the attributes of aparticular destination of which they have noprevious experience (e.g. Um and Crompton,1990). It has also been suggested that previ-ous experience leads to a much more diversi-fied and detailed demand for informationand level of awareness (Gyte and Phelps,1989; Watson et al., 1991). Most of the workdiscussed focuses on destination choices,more specifically holiday destination choicesand the activities undertaken during that hol-iday. Only one study (Darnell and Johnson,1999) focused on a single attraction.

Our study differs in two important regardsfrom previous research. First, rather thanexamining a single choice facet, it addressesfull activity patterns. Secondly, rather thanfocusing on holiday destinations, this study isconcerned with a visit to a theme park duringa day trip.

This chapter is organized as follows. First,hypotheses are formulated based on an evalu-ation of previous research. Next, the method-ology used in this study is discussed.Specifically, the sequence alignment methodthat is used to classify the respondents intohomogeneous clusters based on their activitypatterns in the park is described. Then, datacollection procedures will be outlined. This isfollowed by a discussion of the main results ofthe analysis. The article is completed by sum-marizing the study and discussing some issuesfor future research.

Hypotheses

The aim of this study is to compare the activ-ity patterns of first-time and repeat visitors ina theme park environment in order to gaininsight into differences and similarities intheir respective visitation patterns. The exist-ing literature seems to suggest that first-timevisitors are involved in more active and moreintense activities. Thus, our first hypothesiscan be formulated as follows:

H1: First-time visitors differ from repeat visitorsin their activity patterns in a theme park in thesense that first-time visitors spent more time in thepark, choose more activities than repeat visitors andare more inclined to follow the indicated route inthe park.

This hypothesis is based on the notion thatfirst-time visitors have less information aboutthe theme park, and therefore are moreinclined to explore more possibilities ratherthan being more selective in making theirchoices. However, it may very well be that,especially in a theme park, first-time visitorscan use information sources to improve theirknowledge about the destination, implyingthey are better aware of the alternatives whenmaking their activity choices. At least twosources of information are relevant in thiscontext: information about the activities avail-able in the park and information about thespatial layout of the park. If first-time visitorswould indeed use this information, their activ-ity patterns would be more similar to the activ-ity patterns of repeat visitors. This brings us toour second, more elaborated, hypothesis:

104 A.D.A.M. Kemperman et al.


H2: The difference in activity patterns betweenfirst-time and repeat visitors is moderated by the useof information about the activities available in thepark and the use of information about the spatiallayout of the park.

Methodology

In order to test these hypotheses, the follow-ing methodology was applied. First, theactivity patterns of visitors were classifiedusing the sequence alignment method(SAM) recently introduced in tourismresearch (Bargeman et al., 2002). The SAM(Kruskal, 1983), originally developed forcomparing different DNA and RNA struc-tures, is a method to measure the biologicaldistances among such series of alphabeticcharacters. Assume that an activity patterncan be represented as a string of informa-tion. For example, the sequence of activitychoices made by a visitor during a day visitto a park can be represented as a string ofletters, each letter representing a particularactivity type. Consider two sequences to becompared, s (source sequence) and g (targetsequence). Each element of these sequencesor strings contains a particular alphabeticcharacter.

The sequence alignment method definesdissimilarity as the total amount of effortrequired to change sequence s into g byapplying identity, substitution, insertion anddeletion operations. Each operation involves acertain amount of effort. The magnitude ofthis effort is denoted by weight we(si,gj),ws(si,gj), wd(si,φ) and wi(φ,gj) for respectivelyequality (si=gj), substitution (si�gj), deletionand insertion operations. The ‘φ’ symbolimplies that no operation is applied to theelement denoted by φ. Because strings canbe equalized using many different possibletrajectories, an additional operational deci-sion is required to calculate the similaritymeasure. The SAM is based on the calcula-tion of the Levenshtein distance, which isdefined as the smallest sum of operationweighting values to change sequence s intosequence g. The equation for the ‘weighted’Levenshtein distance is:

d(s,g) = d(sm,gn) (1)d(s0,g0) = 0 (2)d(s 0,g j) = d(s 0,g j-1)+wi(�,gj) (3)d(s i,g 0) = d(s i-1,g 0)+wd(si ,�) (4)d(s i,g j) = min [d(s i-1,g j-1)+w(si ,gj), d(s i,g j-1)+wi(�,gj), d(si-1,gj)+wd(si,�)] (5)

with

we(si ,gj) = 0 if si = gjw(si,gj) = (6)ws(si ,gj) > 0 if si � gj

where i, j ≥1; d(s,g) is the total cost of equal-ization of s (= sm) with g (= gn); m and n arethe number of elements in sequences s and g,respectively; d(s i,g j) is the cost of equalizationof s i with g j, cumulated from the equalizationof s 0 to g 0.

The Levenshtein distance can be used as ameasure of dissimilarity between activity pat-terns. Subsequently, conventional clusteranalysis can be used to classify the visitors intohomogeneous groups based on the distancemeasures.

Next, CHAID (chi-squared automaticinteraction detection) can be used to studythe relationship between the dependent vari-able (i.e. the clusters), and a series of predic-tor variables. CHAID finds the explanatoryvariables that optimally predict the clusters.CHAID is a tree classification method origi-nally developed by Kass (1980), and is anevolution of automatic interaction detection(Sonquist and Morgan, 1964), that employs ahierarchical binary splitting algorithm.CHAID has the advantage of allowing multi-ple-way instead of only binary splits of thepredictor variable. CHAID merges those cat-egories of the predictor variable that arehomogeneous with respect to the dependentvariable, but will maintain all categories thatare heterogeneous. CHAID models are pre-sented in the form of a tree, in which eachfinal node represents a group of homoge-neous categories concerning the dependentvariable. CHAID can be applied to nominalor categorized dependent variables. It pro-vides optimal splits by maximizing the signifi-cance of the chi-squared test at each split. Ifthe statistical significance for the respectivepair of predictor categories is significant,then a Bonferroni adjusted P-value for theset of categories for the respective predictoris computed.

Comparing First-time and Repeat Visitors’ Activities 105


In the present study, CHAID analysis isused to test for each predictor variablewhether there is a significant relationshipwith the dependent variable, the clusters ofactivity patterns. As will be discussed later,two of these variables are categorized as com-binations of first-time/repeat visitor andinformation use. The outcomes of a CHAIDanalysis will then tell whether these cate-gories are split or not.

Data

Sample

To collect the data for this study, a ques-tionnaire was administrated among visitorsin a theme park in The Netherlands in thesummer of 1996. The park that was studiedis especially targeted to children. A conve-nience sample of 2094 adults was selected.Respondents were invited to participate inthe survey and, if willing to do so, asked tofill out the survey as soon as possible aftertheir visit to the park. Respondents wereasked to complete the questionnaire as arepresentative of their travel party, whichincluded children. A pre-stamped returnenvelope was provided. A total of 357respondents returned the questionnaire,representing a response rate of 17% (Table9.1).

A large proportion of respondents wasfemale. Most respondents were from amedium- or high-education and -incomegroup. This finding is possibly due to thefact that the theme park has a strong focuson educational elements in the park, andthe park has no ‘hard thrill’ rides. The activ-ities in the park are very child-friendly andthe children are getting actively involved inthe theatres, live entertainment and attrac-tions. Alternatively, more highly educatedpeople might be more likely to respond tothis questionnaire.

The percentages of the number of chil-dren in specific age groups showed clearlythat most children visiting the theme parkwere in the age group from 0 to 10 yearsold. Hardly any children from 11 to 18 years

old visited the park. Therefore, it can beconcluded that the main market segmentsfor this theme park are households withyoung children, and school groups from pri-mary schools.


Table 9.1. Sample characteristics.

Variable %

GenderFemale 69.5Male 30.5

EducationLow 16.0Medium 41.7High 42.4

Number of adults in group1 10.72 65.23 7.64 6.2�5 10.3

Number of children aged 0–5 in group0 41.51 27.22 22.13 4.4�4 4.8

Number of children aged 6 to 10 in group0 36.11 30.62 20.13 3.7�4 9.5

Total visits to the park1 65.2�2 34.8

IncomeLow 33.7Medium 29.4High 36.9

Group size1 person 0.42 persons 4.93 persons 17.94 persons 40.05 persons 14.7�6 persons 22.1

Type of information usedFriends and relatives 12.9Brochures 21.1Theatre agenda 24.1Map 62.9Signs 21.1Information service 7.5Characters in the park 15.0


Survey instrument

Respondents were asked all kinds of ques-tions about their visit to the park, such astime spending in the theme park, their pref-erences for various facilities in the park, thetype of information used in planning theiractivities and their socio-demographics. Toindicate their revealed choice behaviour inthe park, the respondents were asked to fillout their activity patterns during their visit ina table. Specifically, they were asked to indi-cate at what time during the day they visitedthe various activities, if any, and how muchtime they spent on each of the activities. Theavailable activities in the park were presentedin the first column of a table, and a time axisfrom 9.00 a.m. until 6.00 p.m. was presentedin the first row (see Table 9.2 for a genericdescription of the most important activitiesavailable in the park, and Fig. 9.1 for a simplemap of the theme park and the location of

the activities in the park). For reasons of con-fidentiality, the descriptions of the themepark and the activities in the park are onlygiven in generic terms. However, note that inthe questionnaire the activities were alldescribed as the real existing activities knownto the respondents. The respondents couldindicate their time use by drawing a line foreach of the activities they wanted to visit,from the point in time they started walking tothe activity to the point in time they wouldleave the activity. Next, they were asked toindicate the walking and waiting time foreach activity, by converting the single lineinto a double line. Respondents were pro-vided a map of the park to help them in find-ing the location of activities.

Background information

An analysis of the surveys indicated that therewere two types of visitor groups. One groupconsists of the households who visited thepark with three, four or five persons, consist-ing of one or two adults and children. Theother type of visitors consisted of schoolgroups with, of course, a larger group size.

Furthermore, the results showed that34.8% of the visitors had already visited thetheme park once or several times before.This is a relatively high repeat rate com-pared to other tourist attractions in The


Table 9.2. A description of the type of activities.

Activities Description

1 Welcome activity2 Retail outlet3 Fantasy character4 Theatre5 Attraction6 Food outlet7 Fantasy character8 Theatre9 Attraction10 Attraction11 Attraction12 Fantasy character13 Theatre14 Fantasy character15 Attraction16 Theatre17 Attraction18 Theatre19 Fantasy character20 Food outlet21 Food outlet22 Food outlet23 Attraction24 Attraction25 Farewell activity26 Other activities

1

2

3

4

5

6

1887

9 9

9

22

9

231011

2412

13

1615

17

1921

25

Fig. 9.1. A simple map of the park and the locationof the activities. Note: activities 14 and 20 can befound at various locations.


Netherlands (NRIT, 1998). We also askedrespondents in the survey if they were likelyto repeat their visit to the attraction: 71.7%said they would, 22.9% were undecided,while 5.4% did not plan to come back to thepark. This is a positive result for the themepark, as visitors seem to have liked their visitto the park.

The percentages of the types of informationused by the visitors in the park for planningtheir activities indicate that a large group of vis-itors used the map of the park, 62.9%. Notethat all types of information mentioned areprovided to the respondents at the entrance ofthe park or are available to the respondentsduring their visit. The theatre agenda, whichwas also provided at the entrance of the park,was used by 24.1% of the visitors in the themepark. All other types of information were usedby small groups of visitors.

Analysis

The sequence alignment method was used toanalyse and cluster the respondents’ activitypatterns in the theme park. Of the 357returned questionnaires, 294 were useful forpresent study, because the tables with activitychoices made in the park were recorded cor-rectly. The following procedure was used toanalyse the revealed activity choices made bythe respondents.

First, the revealed activity choices of therespondents in the park were converted to asequence format. Each respondent has onesequence string of pairs of activity types andactivity duration. Duration is calculated in min-utes. Every sequence string starts with activity 0and its duration is calculated from 9 a.m.Every sequence string ends with activity 0 andits duration is calculated from the end of thelast activity until 8 p.m. Furthermore, this‘activity 0’ is also inserted when the data have atime gap between the end time of the previousactivity and the start time of the next activity,although this did not happen often. Thelength of the sequence string (= the numberof activities) differs between persons, while thetotal duration (= the summation of durationacross activities) is the same for all respondents(that is from 9 a.m. to 8 p.m.).

Secondly, a uni-dimensional SAM includ-ing the sequences, embedding informationabout both activity type and activity duration,was applied to calculate the distancesbetween all pairs of activity patterns. Theoriginal operation weights of insertion, dele-tion and substitution were weighted by dura-tion as follows.

Identity = 0; Insertion = (Insertion weight)�((Duration ofthe inserted activity)/30); Deletion = (Deletion weight)�((Duration ofthe deleted activity)/30); Substitution = Insertion �((Duration of theinserted activity)/30) + Deletion �((Durationof the deleted activity)/30).

The normalizing factor 30 is from the obser-vation of the mean duration of 27.65 min,and most frequent duration of 30 min ofthe total of 3412 activities (excluding activ-ity 0) chosen by the respondents. Hence,the weight of an activity’s deletion, insertionand substitution is affected by the amountof time of that activity compared to the aver-age duration.

Thirdly, a cluster analysis was run on theresults of the SAM. Ward’s method was usedbecause it is known to result in relativelyhomogeneous clusters. After investigating thedendrogram, a total of seven clusters wasidentified.

In the subsequent analysis, CHAID wasused to investigate if any statistically signifi-cant differences existed between the depen-dent variable ‘cluster membership’ andbetween the predictor variables: (i) first-time/repeat visitors; (ii) a combination offirst-time/repeat visitors and the use of infor-mation sources about the activities availablein the park; and between (iii) first-time/repeat visitors and the use of spatialinformation about the park. The variable‘information about the activities used’ meansthat one or more of the following informa-tion sources were used: (i) theatre agenda;(ii) information provided by fantasy charac-ters; and (iii) information service in the park.The variable ‘spatial information used’ meansthat one or more of the following informa-tion sources were used in planning the activi-ties: (i) signs in the park; and (ii) a map ofthe park. Three CHAID analyses were con-



ducted with the dependent variable and eachpredictor variable one at a time to test thehypotheses as stated in the introduction. Foreach split chi-squared statistic, degrees offreedom and P-value is given to indicatewhether there is a significant effect of thepredictor variable on the clusters.

Results

This section first describes some characteris-tics of the clusters that were identified, usingthe sequence alignment approach. This is fol-lowed by the results of the three CHAIDanalyses that were used to test the hypotheses.The results describe whether there is a rela-tionship between the clusters of activity pat-terns and whether a person is a first-time orrepeat visitor and whether or not certaininformation sources were used in planningthe activities in the park.

A total of seven clusters was identifiedusing the sequential alignment method,which takes into account the sequence andduration of the activities conducted in thepark. The clusters and some of their charac-teristics are described in Tables 9.3 to 9.5.Specifically, Table 9.3 shows the number ofrespondents belonging to each cluster, theaverage number of conducted activities percluster, the average time spent in the parkand the average arrival and departure timesof each cluster. Table 9.4 indicates the aver-age time spent on an activity per cluster.Table 9.5 describes a representative activitypattern for each cluster. These representative

activity patterns are based on the respondentthat has the mean distance to all other mem-bers of the cluster (calculated on the basis ofthe distance matrix). The representative activ-ity patterns are indicated in the simple mapof the park and shown in Figs 9.2–9.8. Notethat only the order in which the activities arefollowed by the visitors is indicated. We donot know the actual route they followed inthe park to arrive at each of the activities.Furthermore, activities 14 and 20 are notindicated on the map because they can befound at various locations, and therefore theyare not included in the activity patterns aspresented in the figures.

The clusters 2–7 are almost of the samesize (between 11.95% (n = 35) and 17.75%(n = 52)), only cluster 1 is a bit smaller thanthe rest, with 7.51% (n = 22) of the 291respondents. Cluster 1 consists of respondentswho spent most time in the park. On average,these respondents spend 417 min in the park,compared to the average of 372 min spent inthe park by the total sample. The averagenumber of activities chosen by the respon-dents in this cluster is 13 compared to 12activities by the total group, although there isquite some variation in the number of activi-ties chosen. The respondents in this clusterspent on average least time per activity, 24min. As graphically presented in Fig. 9.2, therepresentative activity pattern of cluster 1shows a very strict order in activities chosenby the visitor.

The activity patterns of cluster 2 are, withrespect to time use, arrival and departuretimes, and time spent on the activities quite


Table 9.3. The clusters and their characteristics.

Average number of Average time (in min) Average arrival Average departure activities chosen + spent in the park + time (hh:mm) + time (hh:mm) +

Clusters No. resp. (SD) (SD) (SD mm) (SD mm)

1 22 13 (4) 417 (78) 10:27 (42) 17:24 (1:11)2 49 12 (3) 378 (60) 10:28 (31) 16:46 (51)3 35 11 (4) 375 (63) 10:46 (34) 17:00 (45)4 48 10 (4) 355 (67) 10:33 (36) 16:28 (1:06)5 39 10 (3) 330 (66) 11:21 (50) 16:50 (52)6 52 13 (2) 401 (56) 10:19 (24) 17:01 (52)7 49 12 (3) 361 (57) 10:39 (38) 16:41 (47)

Total 294 12 (3) 372 (67) 10:38 (40) 16:49 (56)


average. However, the persons in this clusterspend more time on activities 8, 12, 13, 24 and25 and less time on activities 6 and 26. Thismeans that on one hand they spent less timeat a particular food outlet and the ‘other cate-gory’, and on the other hand they spent moretime on two of the theatres, and the farewellactivity. The representative activity pattern forthis cluster shows a very diverse pattern inactivities chosen throughout the day.

Also, cluster 3 contains respondents whoseem to be quite average in the characteristicsof their activity patterns, although they arriveand leave the park a bit later than the averagevisitor. Also, they choose in total one activityless than the average and therefore, of allclusters, they spent most time per activity.Specifically, activity 9 is the favourite attrac-tion of these respondents. This is an attrac-tion specifically designed for young children.They spent 57 min at this attraction, which is

15 min more than the average. A lessfavoured activity of this cluster is attraction11, which is targeted to children in the olderage group. The activity pattern of the repre-sentative visitor as presented in Fig. 9.4 showsa quite strict order in activities chosen.

Clusters 4 and 5 are, based on the charac-teristics of their activity pattern described inTable 9.3, quite similar. The respondents ofthese clusters spent on average least time inthe park and also chose the smallest numberof activities. A difference is that the membersof cluster 5 arrive really late in the park, at11.21 a.m. There are also differences in theamount of time spent at the particular activi-ties. Cluster 4 on average is less in favour ofactivity 6 (a food outlet), while this clusterprefers activity 11 (an attraction for olderchildren). The respondents in cluster 5 spent,on average, less time at activity 11 (an attrac-tion), while they spent a lot of time at activity


Table 9.4. Average time spent at an activity per cluster.

Activity Clust. 1 Clust. 2 Clust. 3 Clust. 4 Clust. 5 Clust. 6 Clust. 7 Average

1 27 27 26 25 21 27 30 262 15 17 14 16 11 15 16 153 13 16 20 24 20 16 19 184 17 16 18 15 22 18 17 175 15 17 15 14 18 14 15 156 27 15 30 14 25 17 25 227 22 28 32 23 27 27 31 278 18 28 22 19 24 21 22 229 36 40 57 33 35 44 47 4210 20 29 26 27 30 29 19 2611 22 20 10 23 10 24 20 1812 22 36 24 25 29 34 34 2913 26 48 40 36 38 44 37 3814 20 22 28 24 21 26 20 2315 25 19 23 22 25 30 8 2216 48 56 57 49 45 54 54 5217 14 15 14 11 18 11 12 1318 16 25 26 18 20 32 22 2319 20 19 19 21 21 19 19 2020 21 24 31 32 29 28 29 2821 20 26 19 26 26 30 20 2422 26 27 23 23 21 24 30 2523 17 24 20 29 22 24 32 2424 26 34 33 25 26 26 27 2825 21 32 25 20 22 27 24 2426 27 18 21 22 42 20 21 24

Average 24 29 32 26 28 29 29 28


26 (the ‘other’ category, consisting of allkinds of small and diverse activities in thepark). A large difference can be seenbetween the order of activities chosen in thepark, as shown in Figs 9.5 and 9.6. The repre-sentative visitor of cluster 4 shows a verydiverse order in the sequence of conductedactivities, whereas the representative visitor ofcluster 5 follows a very strict order.


Table 9.5. Representative activity patterns percluster.

Time in Activities Start End min.

Cluster 12 11:00 11:20 20

16 12:30 13:15 455 13:30 13:45 159 14:00 15:00 607 15:00 15:25 25

21 15:30 15:50 2024 16:30 17:30 6014 17:30 17:50 20

Cluster 21 10:30 10:50 203 10:50 10:55 5

17 10:55 11:05 104 11:05 11:20 15

16 11:20 12:45 8515 12:45 12:55 1010 12:55 13:30 3523 13:30 13:40 1026 13:40 13:50 109 13:50 15:00 70

13 15:00 15:20 2019 15:20 15:45 2524 15:45 16:45 602 16:45 17:00 15

Cluster 326 11:25 11:35 101 11:35 11:45 10

16 11:45 12:45 605 12:45 12:50 56 12:50 13:00 108 13:00 13:15 15

23 13:15 13:30 1513 13:30 14:30 609 14:30 15:45 75

14 15:45 16:25 4010 16:25 17:15 5025 17:15 17:30 15

Cluster 41 9:30 10:00 30

17 10:15 10:30 154 10:30 10:45 155 10:45 11:00 156 11:00 11:15 15

19 11:15 11:30 1522 11:30 11:45 159 11:45 12:00 15

16 12:00 12:30 3023 12:30 13:00 3013 13:45 14:15 3012 14:45 15:15 3015 15:45 16:00 15

Continued

Table 9.5. Continued.

Time in Activities Start End min.

Cluster 526 11:00 11:15 153 11:15 11:45 30

17 11:45 12:15 304 12:15 12:30 15

16 12:30 13:00 3021 13:00 13:30 309 13:30 13:40 10

13 13:40 14:30 5011 14:30 14:40 1023 14:45 15:00 159 15:00 15:50 50

20 15:50 16:30 40Cluster 6

1 10:30 11:00 3017 11:00 11:10 104 11:15 11:30 155 11:30 12:00 30

16 12:00 13:15 759 13:15 14:10 55

13 14:10 14:45 3519 14:50 15:10 209 15:15 15:30 157 15:30 15:50 20

18 15:50 16:30 4014 16:30 17:00 302 17:00 17:10 10

25 17:15 17:30 15Cluster 7

1 10:00 10:30 3017 10:30 10:40 104 10:40 10:50 105 10:50 11:00 109 11:00 13:15 135

23 13:15 13:40 2513 13:40 14:15 3526 14:15 14:40 2512 14:40 15:00 2026 15:00 15:15 158 15:15 15:45 30



1

2

3

4

5

6

1887

9 9

9

2

9

231011

24

1213

1615

17

1921

25

Cluster 1

Fig. 9.2. A representative activity pattern for cluster 1.

1

2

3

4

5

6

1887

9 9

9

22

9

231011

24

1213

1615

17

1921

25

Cluster 2


1

2

3

4

5

6

1887

9 9

9

22

9

231011

24

1213

1615

17

1921

25

Cluster 3


1

2

3

4

5

6

1887

9 9

9

22

9

231011

24

1213

1615

17

1921

25

Cluster 4


1

2

3

4

5

6

1887

9 9

9

22

9

231011

24

1213

1615

17

1921

25

Cluster 5


1

2

3

4

5

6

1887

9 9

9

22

9

231011

24

1213

1615

17

1921

25

Cluster 6



Like cluster 1, cluster 6 consists ofrespondents who spent more time than aver-age in the park and also chose more activi-ties; however they were less diverse (smallstandard deviation). The arrival and depar-ture times were both a bit earlier than forcluster 1. Contrary to cluster 1, the respon-dents in cluster 6 spent about the averagetime per activity. Furthermore, they likedmore than average activities 16 and 18, anattraction and a theatre, respectively. Also,the representative activity pattern of thiscluster is much more diverse than the activ-ity pattern of cluster 1.

Finally, cluster 7 seems to be average withrespect to the characteristics of the activitypatterns presented in Table 9.3, although dif-ferences can be seen for the time spent atparticular activities. The members of this clus-ter do not like activity 15 (an attraction), atall, spending only on average 8 min at thisattraction. Also, they spent less time on activ-ity 10 (another attraction). On the otherhand, they spent more time on average onactivities 9 (an attraction), 12 (a fantasy char-acter), 22 (a food outlet) and 23 (an attrac-tion). Figure 9.8 shows a representativeactivity pattern for cluster 7. The visitor had aquite strict order in the activity choices madeduring his or her visit to the park.

CHAID analysis was used to investigatewhether any statistically significant differ-ences existed between the dependent variable‘cluster membership’ and between the pre-

dictor variables: (i) first-time/repeat visitors;(ii) a combination of first-time/repeat visitorsand the use of information sources about theactivities available in the park; and between(iii) first-time/repeat visitors and the use ofspatial information about the park.

The results of the CHAID analyses are pre-sented in Figs 9.9–9.11. On top of the treediagram in each figure, the distribution ofthe respondents over the dependent variable,the clusters of activity patterns, is given.Below this square the predictor variable isgiven with the statistics for the split; the P-value, the chi-squared statistic and thedegrees of freedom. Then the split is indi-cated with above each branch thecategory(ies) of the predictor variable thatis(are) homogeneous with respect to thedependent variable. For each (group of) pre-dictor category(ies) the number and percent-age of respondents per cluster are given.

Figure 9.9 presents the results of the firstCHAID analysis between the clusters of activ-ity patterns and the predictor variable first-time/repeat visitors. Note that this is similarto a chi-squared test. The results indicatethat there is a small significant (P = 0.0773)effect of first-time visitors and repeat visitorson the clusters of activity patterns. Thismeans that there is a difference in activitypatterns between persons that visit the parkfor the first time and repeat visitors. Theresults in the figure indicate that first-timevisitors were dominant in clusters 1 and 5and that repeat visitors are more dominantin clusters 4 and 6.

Specifically, this means that there is agroup of first-time visitors, cluster 1, that stayslonger in the park than average, and choosesa larger number of activities than the averagevisitor. This would be an indication of sup-port for the first hypothesis: first-time visitorschoose on average more activities in theiractivity pattern and spend more time in thepark. However, first-time visitors in cluster 5spent the least time in the park, and chosethe least activities. This would not support thefirst hypothesis. However, the members ofcluster 5 spent a lot of time on the ‘other’ cat-egory. In that sense, it would support thestatement that first-time visitors choose notonly the best known sites.


1

23

4

5

6

1887

9 9

9

22

9

231011

24

1213

1615

17

1921

25

Cluster 7



For the repeaters, the same contradictioncan be seen with regard to the number ofactivities chosen and the time spent in thepark as for the first-timers. The repeat visitorsin cluster 4 spent almost the least time in thepark and chose the least number of activities,while repeaters in cluster 6 spent almost themost time in the park and chose the mostnumber of activities.

When comparing the order in activitychoices made by the visitors in the variousclusters, an interesting result can be seen. Thefirst-time visitors, both in cluster 1 and 5, fol-low a very strict route in the park, while therepeat visitors in cluster 4 and 6 show a verydiverse order in activities chosen during theirvisit in the park. Thus, based on this empiricalresult, hypothesis 1 can be supported.

It can be concluded that there is a signifi-cant difference between first-time and repeatvisitors based on their activity patterns in thepark. The most important difference is that

first-timers strictly follow the route as indi-cated in the park, while repeat visitors followtheir own, diverse route.

In the introduction to this chapter weassumed that first-timers will be closer torepeaters in their activity pattern when theyuse information sources to improve theirknowledge about the park to account fortheir lack of experience with the park.Specifically, we defined two new variables withfour categories each. The first variable wasdefined as whether a respondent was a first-time or repeat visitor and whether informa-tion sources about the activities in the park(information service in the park, the theatreagenda and the fantasy characters in thepark) were used or not used. The second vari-able was defined as whether a respondent wasa first-time or repeat visitor, and whether spa-tial information sources about the park (mapof the park and signs in the park) were usedor not used.


First-time/repeat visitorsP-value=0.0773

Chi-squared=11.3814d.f .=6

Cluster1234567

Total

%7.51

16.7211.9516.3813.3117.7516.38(100)

n22493548395248

293

Cluster1234567

Total

%9.95

15.1813.0914.6615.1814.6617.28

(65.19)

n19292528292833

191

Cluster1234567

Total

%2.94

19.619.80

19.619.8

23.5314.71

(34.81)

n3

201020102415

102

First-time visitors Repeat visitors

Clusters

Fig. 9.9. Results of CHAID based on the dependent variable clusters and the predictor variable first-time/repeat visitors.


The results of the second CHAID analysisbetween the clusters of activity patterns andthe predictor variable first-time/repeat visi-tors combined with information (not) usedabout the activities are shown in Fig. 9.10.

The results show that there is a significanteffect of the predictor variable on the clustermembership (P = 0.0249). The two branchesindicate that there is a distinction betweenfirst-time and repeat visitors that do use infor-mation about the activities in planning theiractivity pattern, and first-time and repeat visi-tors that do not use information about theactivities available in the park. The first twocategories are more dominant in clusters 3and 6, while the other two categories aremore present in clusters 1 and 2.

A remarkable result is that a difference inactivity patterns can be seen betweenwhether the information about the activitieswas used or not used, and not between first-

time and repeat visitors. These findings cor-respond partly with hypothesis 2. First-timevisitors that use information about the activi-ties differ with respect to their activity pat-terns from first-time visitors that do not usethese information sources. However, repeatvisitors that use information about the activi-ties are different, with respect to their activ-ity patterns, from repeat visitors that do notuse information. An explanation for thismight be that the information about theactivities differs per season and thereforeper visit, and it thus seems difficult todevelop knowledge about the activities inthe park. This causes the difference betweenvisitors that do use this type of informationin planning their activities and visitors thatdo not use this information.

The results of the third CHAID analysis,between the clusters of activity patterns andthe predictor variable first-time/repeat visi-


First-time/repeat visitorsand information used about activities

P-value=0.0249Chi-squared=14.4569

d.f.=6

Cluster1234567

Total

%7.51

16.7211.9516.3813.3117.7516.38(100)

n22493548395248

293

Cluster1234567

Total

%12.9629.635.5611.119.26

12.9618.52

(18.43)

n7

163657

1054

First-time + info usedRepeat + info used

First time + information not usedRepeat + information not used

Clusters

Cluster1234567

Total

%6.28

13.8113.3917.5714.2318.8315.90

(81.57)

n15333242344538

239

Fig. 9.10. Results of CHAID based on the dependent variable clusters and the predictor variable first-time/repeatvisitors and use of information about the activities.


tors and spatial information (not) used arepresented in Fig. 9.11. A significant differ-ence (P = 0.0152) can be noticed – first-timevisitors that use spatial information differfrom first-time and repeat visitors that do notuse this information, and also from repeat vis-itors that use this type of information. Thefirst-time visitors that use the spatial informa-tion are dominant in clusters 1 and 3, whilethe other categories are dominant in clusters4 and 6.

This finding partly confirms the validity ofhypothesis 2. First-time visitors that use spatialinformation about the park differ withrespect to their activity patterns from first-time visitors that do not use informationsources. Furthermore, repeat visitors that usespatial information about the park do not dif-fer with respect to their activity patterns thanrepeat visitors that do not use information.

However, it is remarkable that first-time visi-tors that use spatial information differ fromrepeat visitors that use or do not use thisinformation source.

It seems that first-time visitors that usespatial information (a map of the park andsigns in the park) follow the route in thepark as indicated by the management of thepark, while the others that do not use thatinformation, or use it, still follow a morediverse route in the park. An explanationfor this result might be that first-time visi-tors that are aware of the layout of the parkuse the park in that sense, and the othergroups seem to have a more ad hoc, diverseactivity pattern in the park. However, repeatvisitors that use information may use itspecifically to visit particular activities, butthis does not have to take place in the orderindicated.


First-time/repeat visitorsand spatial information used

P-value=0.0152Chi-squared=15.7509

d.f.=6

Cluster1234567

Total

%7.51

16.7211.9516.3813.3117.7516.38(100)

n22493548395248

293

Cluster1234567

Total

%4.32

19.149.26

20.3711.7320.9914.2

(55.29)

n7

311533193423

162

First-time + info used First time + information not usedRepeat + information used

Repeat + information not used

Clusters

Cluster1234567

Total

%11.4513.7415.2711.4515.2713.7419.08

(44.71)

n15182015201825131

Fig. 9.11. Results of CHAID based on the dependent variable clusters and the predictor variable first-time/repeat visitors and use of spatial information.


Conclusion and Discussion

The aim of this study has been to comparethe activity pattern choices of first-time andrepeat visitors. As such, the results of thisstudy contribute to the literature on first-time versus repeat visitor behaviour. In par-ticular, it differs from previous researchefforts in that the focus of research is oncomprehensive, multi-dimensional activitypatterns as opposed to particular isolatedfacets of such patterns, and in that the focusis on a theme park as opposed to holidayexperiences. The study also attempts to studycomplex, interrelated choices of tourists.

The guiding hypothesis underlying theanalysis was that activity patterns of first-timevisitors tend to differentiate from the activitypatterns of repeat visitors, mediated by theiruse of information. Differences between thetwo groups were assumed to be reduced whenfirst-time visitors use information about theavailable activities and the spatial layout ofthe theme park.

The results tend to partly support thishypothesis. First, the results of the analysisindicate that first-time and repeat visitors dif-fer with respect to their activity patterns inthe park. The most important difference isthat first-time visitors follow a very strictroute in the park as indicated by theme parkmanagement, while repeat visitors have amore diverse activity pattern. With respect tothe number of activities chosen and the timespent in the park, it can be concluded that,compared to the average, there is a clusterof first-timers that spent a long time in thepark and choose a large number of activities;however, there is also a cluster that spent lesstime in the park and chose less activitiesthan on average. This also applies for therepeat visitors.

Secondly, it can be concluded that thedifference between the two groups isreduced when first-time visitors use informa-tion about the available activities and thespatial layout of the park. Specifically, a dif-ference in activity patterns was foundbetween visitors that use information aboutthe activities and visitors that do not use thistype of information, no matter whether they

were first-time or repeat visitors. It seemsthat information about the activities mightchange over time, for example per season,and therefore, repeat visitors do not havethis knowledge of previous experiences inthe park. Also, an effect of the use of spatialinformation about the park on visitors’activity pattern can be observed. However, itseems that spatial information is more con-stant over time.

Overall, it seems that both hypotheses aremost strongly confirmed as far as thesequence of activities is concerned. It shouldbe recalled that the very reason for applyingthe sequence alignment method is that it par-ticularly captures the sequence of conductedactivities. Hence, the identified clusters aremost homogeneous in terms of the activitysequence. The number of activities and dura-tion are more or less derived. Hence, part ofthese results may be specific for the appliedmethodology. More detailed specific analysescan be conducted to further elaborate thepresent findings, but are not reported heredue to size limitations.

The relevance of this finding lies not onlyin its elaboration of our current knowledge. Amore thorough comprehension of visitors’preferences for different activity patterns in atourism environment is also highly relevantbecause it is an important component for anyeffective planning, management and market-ing of the tourism attraction. Site planningsets the components for the visitor experi-ence. Therefore, control of visitor use andflows is a basic consideration in much visitorattraction planning. Establishing carryingcapacities of attractions and applying tech-niques to organize visitor flows and to controlover-usage is an important factor. The resultsof the present study, in addition to provinginformation about typical activity patterns,also suggest that the right provision of infor-mation may induce visitors to be more selec-tive, offering a means of better organizingvisitor flows.

From a methodological perspective, thisstudy has provided additional empirical evi-dence about the potential usefulness of theSAM to classify tourists in terms of their overtbehaviour. Yet, the present study only used auni-dimensional measure of sequence



alignment. The reason for this choice is thatit is most appropriate in the present case.However, for other types of behaviour, wherethe transport mode, destination and other

facets of activity patterns may also differ, thenewly suggested multi-dimensional SAM (Johet al., 2001, 2002) might be a better candi-date.


References

Bargeman, B., Joh, C.H. and Timmermans, H.J.P. (2002) A typology of tourist vacation behavior using asequence alignment. Annals of Tourism Research 29, 320–337.

Crouch, G.I. and Louviere, J.J. (2000) A review of choice modelling research in tourism, hospitality andleisure. Paper presented at the Second Symposium on Consumer Psychology of Tourism, Hospitalityand Tourism, Vienna, Austria, July.

Darnell, A.C. and Johnson, P.S. (2001) Repeat visits to attractions: a preliminary economic analysis. TourismManagement 22, 119–216.

Dellaert, B.G.C., Ettema, D.F. and Lindh, C. (1998a) Multi-faceted tourist travel decisions: a constraint-based conceptual framework to describe tourists’ sequential choices of travel components. TourismManagement 19, 313–320.

Dellaert, B.G.C., Prodigalidad, M. and Louviere, J.J. (1998b) Family members’ projections of each other’spreference and influence: a two-stage conjoint approach. Marketing Letters 9, 135–145.

Fakeye, P.C. and Crompton, J.I. (1991) Image differences between prospective, first-time and repeat visi-tors to the Lower Rio Grande Valley. Journal of Travel Research 30, 10–15.

Gitelson, R.J. and Crompton, J.L. (1984) Insights into repeat vacation phenomenon. Annals of TourismResearch 11, 199–217.

Gyte, D.M. and Phelps, A. (1989) Pattern of destination repeat business: British tourists in Mallorca, Spain.Journal of Travel Research 28, 24–28.

Jeng, J.M. and Fesenmaier, D.R. (1998) Destination compatibility in multidestination pleasure travel.Tourism Analysis 3, 77–87.

Joh, C.H., Arentze, T.A. and Timmermans, H.J.P. (2001) A position-sensitive sequence-alignment methodillustrated for space–time activity-diary data. Environment and Planning A 33, 313–338.

Joh, C.H., Arentze, A., Hofman, F. and Timmermans, H.J.P. (2002) Activity pattern similarity: a multidi-mensional sequence alignment method. Transportation Research Part B 36, 385–403.

Kass, G. (1980) An exploratory technique for investigating large quantities of categorical data. AppliedStatistics 29, 119–127.

Kemperman, A.D.A.M., Borgers, A.W.J., Oppewal, H. and Timmermans, H.J.P. (2000) Consumer choice oftheme parks: a conjoint choice model of seasonality effects and variety seeking behavior. LeisureSciences 22, 1–18.

Kemperman, A.D.A.M., Borgers, A.W.J. and Timmermans, H.J.P. (2002) A semi-parametric hazard modelof activity timing and sequencing decisions during visits to theme parks using experimental designdata. Tourism Analysis 7, 1–13.

Kruskal, J.B. (1983) An overview of sequence comparison. In: Sankoff, D. and Kruskal, J.B. (eds) TimeWarps, String Edits, and Macromolecules: The Theory and Practice of Sequence Comparison. Addison-WesleyLondon, pp. 1–44.

Louviere, J.J. and Timmermans, H.J.P. (1990) Stated preference and choice models applied to recreationresearch: a review. Leisure Sciences 12, 9–32.

Murphy, P.E. and Pritchard, M.P. (1997) Destination price-value perceptions: an examination of origin andseasonal influences. Journal of Travel Research 35, 16–22.

NRIT (Netherlands Research Institute for Recreation and Tourism, Nederlands Research Instituut voorRecreatie en Toerisme) (1998) Trendrapport Toerisme 1997/’98. NRIT, Breda.

Oppermann, M. (1996) Visitation of tourism attractions and tourist expenditure patterns–repeat versusfirst-time visitors. Asia Pacific Journal of Tourism Research 1, 61–68.

Oppermann, M. (1997) First-time and repeat visitors to New Zealand. Tourism Management 18, 177–181.Oppermann, M. (1998) Destination threshold potential and the law of repeat visitation. Journal of Travel

Research 37, 131–137.Reid, L.J. and Reid, S.D. (1993) Communicating tourism supplier services: building repeat visitor relation-

ships. Journal of Travel and Tourism Marketing 2, 3–20.


Siderlis, C. and Moore, R.E. (1998) Recreation demand and the influence of site preference variables.Journal of Leisure Research 30, 301–318.

Sonquist, J.N. and Morgan, J.A. (1964) The Detection of Interaction Effects, Monograph 35. Survey ResearchCenter, Institute for Social Research, University of Michigan, Michigan.

Taplin, J.H.E. and McGinley, C. (2000) A linear program to model daily car touring choices. Annals ofTourism Research 27, 451–467.

Tideswell, C. and Faulkner, B. (1999) Multidestination travel patterns of international visitors toQueensland. Journal of Travel Research 37, 364–374.

Um, S. and Crompton, J.L. (1990) Attitude determinants in tourism destination choice. Annals of TourismResearch 17, 432–448.

Uysal, M., Fesenmaier, D.R. and O’Leary, J.T. (1994) Geographic and seasonal variation in the concentra-tion of travel in the United States. Journal of Travel Research 33, 61–64.

Watson, A.E., Roggenbuck, J.W. and Williams, D.R. (1991) The influence of past experience on wildernesschoice. Journal of Leisure Research, 23, 21–36.



Chapter tenA Study of Tourist Decision Processes: Algarve, Portugal

Antónia Correia1 and Geoffrey I. Crouch2

1Faculty of Economics,University of Algarve, Campus de Gambelas, 8000-117 Faro, Portugal;2School of Business, La Trobe University, Melbourne, Victoria 3086, Australia

Abstract

The issue of vacation decision making has become a focal point in tourism research during the past twodecades. Despite the increasing amount of published research in this area, practical studies that test theo-ries are very scarce, especially for Portugal. This study investigates tourist consumer behaviour in theAlgarve, with particular emphasis on their perceptions and motivations. A theoretical framework is devel-oped, which assumes a relationship between pre-decision, decision and post-purchasing evaluation, i.e. themain phases of the tourist decision process. A proposed decision model substantiates the motivations, per-ceptions, expectations and choice about the Algarve. A questionnaire was used to interview a random sam-ple of tourists. Multivariate statistics, especially factor analysis, were employed to find different exogenousvariables at work for both perceptions and motivations, which varied according to the participant’s countryof origin. Eight factors were found which help in understanding the perceived image and motivationstourists of different nationalities have about the Algarve. Using univariate analysis, the study went a step fur-ther to characterize the different phases of tourist behaviour. It was found that the Algarve is perceived pri-marily as a sports destination. The sun and the beach also predominate as leading motivations for tourists.In terms of evaluation, all tourists seem to be satisfied, although Germans reveal some dissatisfaction. Theimplications of these findings for explaining consumer behaviour indicate future lines of research.

Introduction

This study pursued two objectives. First, vaca-tion decision-making processes were reviewedin order to propose an empirically testablemodel. The inclusion of perceptions andmotivations within this model is a newattempt to analyse choice in tourism. Second,we explored the usefulness of a model that

extended the decision choice process instudying tourist behaviour in the Algarve.

By initially analysing the usefulness of anextended decision-making process, this studydeveloped a theoretical framework for study-ing this sequenced choice. This is followedby the methodology used to collect the infor-mation, and a characterization of the deci-sion process involved for specific market



countries, e.g. motivations, perceptions,expectations and choice. The implications ofthese findings for explaining consumer deci-sions and the limitations of the study,together with possible avenues for futureresearch, are then discussed.

Theoretical Framework of the TouristDecision-making Process

Most studies on consumer behaviour referto five stages in the decision-making process;namely: (i) identification of needs; (ii)information gathering; (iii) evaluation ofalternatives; (iv) process of choice; and (v)post-purchase processing (Bentler andSpeckart, 1979; Moutinho, 1982; Um andCrompton, 1990; Crompton, 1992;Crompton and Ankomah, 1993; Middleton,1994; Ryan, 1994; Solomon, 1996). Thetourist decision making process adopted inthis chapter presupposes the above-men-tioned stages placed into three majorphases, whose objectives and constraints,when defined, allow optimized behaviour in

the selection of available options. Based onprevious work by Correia (2000), Fig. 10.1seeks to schematize the decision processwhere three essential phases occur: (i) pre-decision; (ii) process of decision; and (iii)post-purchase evaluation.

This figure attempts to define the behav-iour of consumers in terms of a sequenced,organized and dynamic process, where sev-eral factors compete towards the develop-ment of each of the three phases. The first,or pre-decision phase, relies on such condi-tions as predisposition to external stimuli,communication access between the con-sumer and the main sources of information,and ultimately, consumer learning. By learn-ing, we mean the way the consumer receivesthe information and filters what is relevantin order to develop a list of preferredoptions. In this phase, consideration is givento antecedent factors that affect one’s traveldecision, such as perception, motivation andpreference. The second, or the decisionphase, assumes a consideration of both con-sumer income and time available. The con-sumer tourist bases his or her decision

122 A. Correia and G.I. Crouch

Inputs Communication means Filters Motivations OutputsPreferences

Perceiving risk

ChoiceAttitudes

NeedsDesires

Aims

Psychographiccharacteristics

Social andeconomiclevel

Learning

Perceptions

Experience

Sensibilitypost-purchase

Set oftourism

products

Publicitypromotionbrochures

FriendsFamilyGroups

1Process of

pre-decision

2Process ofdecision

3Post-purchase

evaluation Probability ofrepeatingpurchase

Post-purchase information

Expectations

Reality

Confirmed Unconfirmed

Fig. 10.1. A model of tourist’s decision processes.


according to the array of options that fallwithin his or her ideal budget range fortravel and lodging expenses. The third, theevaluation of the post-purchase phase, is theconsequence of other stimuli that influencethe choice process and presupposes con-sumer satisfaction with the undertaken desti-nation. This phase is important in evaluatingthe varying probabilities of repeating thepurchase to that particular destination. It iscommonly found, however, that despiteacknowledging high levels of satisfactionupon termination of travel, consumers donot always repeat the same purchase becausethey want to discover new destinations. Thispremise is legitimate for those market seg-ments known as ‘risk-takers’, those who pre-fer new destinations, in search of othernovel experiences. More cautious tourists,on the other hand, are more likely to returnto familiar destinations.

The issue of vacation decision making hasbecome a focal point in tourism research dur-ing the past two decades. An extensive reviewof destination demand modelling literaturecan be found in Crouch (1994) and Lim(1997). Despite the increasing amount ofpublished research in this area, practical stud-ies that test theories are very scarce, especiallyfor Portugal. Amongst the papers that haveanalysed tourism in Portugal are Syriopoulosand Sinclair (1993), who included Portugalin their AIDS (Almost Ideal Demand System)analysis. Syriopoulos (1995) also included thecountry in an error-correction model. Otherresearch results pertaining to Portuguesetourism demand have arisen from Silva(1991, 1995), Correia (2000) and Matos(2000). Silva (1991) analysed Portuguesetourism demand from the point of view ofcountry-of-origin, with a log-linear model thatis a function of price, income, promotionalexpenditure and the cost of transport fromfive origin countries. Correia (2000) analysedtourism demand for the Algarve region bycountry-of-origin with the aid of two models:the first using data from a questionnaire andthe second using monthly data of five coun-tries of origin for the period 1989 to 1998 inorder to estimate a translog demand equationfor each country. Matos (2000) analysed theaggregate Portuguese tourism demand by

country-of-origin, using eight countries from1977 to 1997. A general characteristic of thispast research into tourism to the Algarve is itsreliance on aggregate data, thus disregardingthe heterogeneity of tourist demand. Theaforementioned literature was based on eco-nomic models, and all the authors sought toexplain tourism demand based upon eco-nomic variables.

In this chapter, using factor analysis weexplain the main driving motivations and per-ceptions of holidaymakers of different nation-alities and different socio-economic profiles.Using univariate analysis, the study also char-acterizes the different phases of tourist behav-iour.

Contextual Setting: Tourism in theAlgarve

According to the WTO (1999), in terms ofinternational tourism receipts, Portugal wasranked 19th as a tourist destination in 1990and 24th in 1998, representing 1.1% of totalglobal tourism demand in 1998. This valuecompares with the figure of 6.7% for itsneighbour, Spain. The Portuguese tourismindustry contributed 3.2% to national grossadded value and employed 5.2% of the activepopulation in 1998. It developed into animportant tourist destination in the 1960s,based mainly on the attraction of both itsbeaches and climate. Representing only 6%of the total area of Portugal, the Algarveranks as the country’s prime tourist destina-tion, attracting 42% of all tourism activity inthe country.

As far as the nationality of the tourists whovisit the Algarve is concerned, we can seefrom Table 10.1 that, in 1989, the main coun-try of origin was the UK, followed byGermany, The Netherlands and Spain. Thisprofile has meant that tourism has becomethe leading economic activity in the region.Moreover, the performance of Algarvetourism is significantly dependent on theEuropean economy in general, and on theUK, Germany and, to a lesser extent, TheNetherlands and Spain in particular. Theregion competes directly with the SpanishMediterranean beach-and-sun market.

A Study of Tourism Decision Processes 123


By considering the aforementioned phasesof the tourist decision-making process andaccepting that the Algarve has three main mar-kets, the UK, Germany and The Netherlands(DGT, 1998a, p. 23), a methodology was devel-oped to analyse the consumer behaviour oftourists arriving from these countries and alsofrom its neighbour, Spain.

Methodology

Characterizing different markets is not aneasy undertaking given their heterogeneityand a general lack of statistical information. Adetailed questionnaire was produced in orderto collect data such as general backgroundinformation, motivation arousal, externalstimuli (reference groups) and internal stim-uli (previous experience, social/economicprofile, and psychographic characteristics),perception formation, and decision con-straints (time and budget). Informationrelated to incoming tourists was also collectedthrough personal interviews with theEuropean Travel Monitor (ETM). The find-ings were limited to those visiting the Algarveduring July and August 1997. Data wasarranged according to country-of-origin.Based on a significance level of 5%, a mini-mum of 384 responses were required. Thequestionnaire consisted of closed questionsthat sought to measure satisfaction levelsrelating to perception, motivation and post-purchase effects, using a five-point Likertscale (1 = not important; 5 = very important).The questionnaires were individually reviewed,codified, issued a magnetic identification tagand administered according to a set schema.Results were processed and handled usingSPSS v.10.

In order to obtain a simplified and inter-pretable representation of the main motiva-tional factors and consumer expectations forthe case of the Algarve, principal compo-nents analysis (PCA) with a Varimax rotationwas applied to those questions related tomotivation and perception. The Kaiser crite-rion was adopted for factor extraction, whichconsisted of excluding components witheigenvalues whose averages were below 1.

The Tourist Decision-making Process

The empirical study was carried out by meansof the previously mentioned questionnaire,which was administered to a stratified, ran-dom sample of tourists at Faro airport(Algarve) between July and August 1997. Thecentral aim was to determine their socio-economic characteristics as tourist consumerswho had chosen the Algarve as their destina-tion. The first phase of the study was a pilotsurvey used to identify the sample populationand clarify certain questions on the question-naire. The final questionnaire was a revisedversion based on the results of the pilot sur-vey. The sample was stratified by country-of-origin according to the parameters of touristvisits published as official statistics. The sur-vey was conducted in the departure area ofthe airport, just before passengers left theregion. The interviews were given in one ofthree languages: English, Spanish or French.This method has an advantage over site-specific interviews in that it allows us to inter-view both those who are travelling fortouristic purposes and those who are travel-ling for other reasons. The random routeprocedure was chosen for sampling. Themaximum number of surveys per day and per


Table 10.1. Tourist visitors to the Algarve.

Principal origin countries 1989 % 1998 %

1. UK 482,424 32 618,480 272. Germany 151,743 10 385,870 173. Netherlands 110,256 7 146,443 64. Spain 71,449 4 76,394 3

Total 1,473,698 100 2,225,000 100

Source: INE (National Statistical Institute).


interviewer was set at ten. Since the surveywas directed at individuals, rather than thefamily, only individuals over 18 years of agewere interviewed. Due to budgetary con-straints and limited time available, theresearch set out to collect data from 500tourists and resulted in 454 usable responses.The distribution of the sample by nationalityis presented in Table 10.2.

The distribution of the sample by country-of-origin closely follows the distribution oftourist numbers as illustrated in Table 10.1. Itcan therefore be asserted that the 454 com-pleted survey responses approximates a strati-fied random sample of tourists in the Algarve.The general characteristics of these respon-dents vary according to nationality, introduc-ing heterogeneity into the sample. Findingsare presented under the three-phase touristdecision-making framework (i.e. pre-decision,decision and post-purchasing evaluation).

Pre-decision phase

The pre-decision phase is triggered by certainconsumer needs which develop or build upinto a subsequent desire to learn more abouta particular destination aimed at satisfying hisor her preference profile, and confirmingfavourable recommendations from others interms of any word-of-mouth. This culminateswith the formation of more specific motiva-tions and perceptions.

The arousal of these needs necessarilydepends on the access consumers have toinformation of common tourist destinations.His or her preferences, socio-economic pro-file, and reference group, condition the way a

tourist learns about a particular destination.Table 10.3 summarizes the main features ofthe pre-decision phase according to the mainorigin markets. These features include desti-nation awareness and preference, and inter-nal and external factors that influencedestination choice.

The Dutch market holds the smallest pro-portion of tourists travelling to theMediterranean, while for France, Germany,Belgium and Luxembourg the Mediterraneanpossesses greater appeal. Spaniards are con-sidered traditional in terms of vacation habitsrestricting most travel to neighbouring coun-tries such as France and Portugal, due toeither geographical proximity or cultural simi-larities. The British tend to prefer France asan ideal travel destination, although Spainmaintains a leading position. The strength ofthe pound and the lowering of airfares, lodg-ing and fuel prices allow the British to enjoythese destinations for longer periods and atmore competitive prices. As a result of reunifi-cation, Germany suffered a severe economiccrisis in the 1990s. The second half of the1990s, however, showed signs of recovery,allowing more Germans to extend their travelto destinations outside of the country. In fact,a growing interest for vacations abroad is adirect result of cuts in airfares and an increasein the number of airlines. A better knowledgeof foreign languages and a wider range andawareness of destination options has stimu-lated more people to travel internationallywho might have otherwise tended to traveldomestically. Outside Germany, the favouritedestinations now are Austria, Spain and Italy.France, Greece and Switzerland are also popu-lar destinations for the Germans, althoughthe number of visits to these countries hasdecreased since 1995.

From an initial consideration set of destina-tions, tourists form an evoked set of destinationalternatives (Woodside and Sherrell, 1977),for which they then gather more specificinformation. Important sources of informa-tion are usually tourist brochures, recommen-dations from friends and family and previousexperience. When considering nearby desti-nations, advice from friends and family isoften sufficient to persuade the consumer inhis or her final decision.


Table 10.2. Distribution of the sample by country-of-origin.

Country Number %

Germany 97 19Netherlands 32 6Spain 32 6UK 200 40Other 93 19Invalid responses 46 9

Total 500 100


Europeans, in terms of environmentalpreference, according to a study carried outby the General Directorate of Tourism(DGT, 1998a), show greater propensity forsun and beach environments, followed bymountains, and then cities. Preference for acertain type of environment derives from theconsumer’s personal motivation. TheGermans and the Dutch continue to show anoverall inclination for the sun and thebeach. Germans, however, tend to value amore recreational environment, such assports, countryside and mountains. TheSpanish reveal a greater preference for vaca-tioning in cities, followed by the traditionalattractions of the sun, the beach and sight-seeing. In the case of the British, leisure aswell as the sun and beach typically deter-mine their destination choices. Sports andsightseeing also constitute attractive featuresfor the younger population. British residentsoften take advantage of the growing numberof special offers for shorter trips abroad.

The tourist’s socio-economic profile influ-ences how motivation and preference evolve.In fact, variables such as residential area,social class, age group and size of householdconstitute important determinants of thefinal choice.

Tourists that visit the Algarve derivemainly from populations who live in cities,except for the British for whom travel toPortugal is relatively evenly spread among res-idents of villages, towns and cities. Inboundtravel to the Algarve is essentially middle- andupper-class tourism. In terms of age group,the largest population percentage that travelsto the Algarve lies in the 30–44 age range, infamily groups of three to four people. This isan indication that family tourism tends topredominate over other types.

External factors were also measured by thequestionnaire. Among these factors were rec-ommendations made by friends and familymembers, previous experiences and theprospect of returning to the Algarve.


Table 10.3. Tourist pre-decision factors according to main visitor countries.

Germany Spain Netherlands UK

Destination Germany, France, Germany, Belgium, Spain, France, preferences Spain, Italy, Austria Portugal Luxembourg, France USA

Information gathering Brochures Brochures, family Brochures Brochuresand friends

Internal factorsTypes of Beach, mountains Beach, city Beach, countryside Beach, surroundings mountains

Purpose of travel Vacation Vacation Vacation VacationTypes of leisure trips Sun and beach, Sun and beach, Sun and beach, Sun and beach,

sightseeing, city sightseeing, city sightseeing, city sightseeing, city

Socio-economic profileResidential area City City City City, villageSocial level Middle/high Middle/high Middle/high Middle/highAge group 30–59 years 15–44 years 15–44 years 15–44 years Household 2–4 persons 2–4 persons 1–2 persons 1–2 persons

External stimuliAverage rate of Not too certain Indefinite approval Indefinite approval Almost certain approval from (3.67) (4.2) (4.25) (4.64)friends

Previous experience 60% new customers, 76% repeat customers, 50% new customers, 63% new customers, 40% repeat customers 24% new customers 50% repeat customers 37% repeat customers

Frequency of None Frequent 1–2 times 1–2 timesrecurrent visits to Algarve

Source: DGXXIII (1997, in DGT 1998b), ETM (1998a–d), Correia (2000).


The choice in most countries, except forGermany, is strongly influenced by relativesand friends. The rate of return to the Algarveis usually evidence of the level of satisfactionof a past visit. In general, the Algarve touristmarket experiences repeat visits by thoseGerman tourists who still regard the region asholding some degree of novelty for them.From Table 10.3, it can be observed that thesame tourist will visit the Algarve on averageonce or twice during his or her lifetime.Accessibility and proximity explain the fre-quent visits of the Spanish.

Motivation arousal

Once informed about a particular destination,this information is stored and understood,prompted by the individual’s interest and moti-vation. For the purpose of this investigation, 14variables, or determinants, were consideredunder PCA, resulting in four motivational fac-tors that explain 55% of the total variation.

Table 10.4 presents the results of the PCAafter Varimax rotation. The KMO test pre-sents a value, which lets us conclude that theextraction is an average one. Moreover, theKaiser test allows us to reject the hypothesisthat any variable is a linear combination ofthe remaining ones.

After Varimax rotation, variables whoseeigenvalues were superior to 0.5 were

divided into four factors. A driving motiva-tion identified as adventure and sport isdefined by determinants of a proactivenature such as ‘to challenge my abilities’, ‘touse my imagination’ and ‘to practise sports’.Knowledge, on the other hand, considers themore social determinants, containing suchvariables as ‘to discover new places andthings’, and ‘to be in a social surrounding’.A third driving motivation, escape/socializa-tion, was found to encompass the compo-nents ‘to avoid day-by-day stress’ and ‘tomake friends’, while leisure and relaxationhighlights the need for physical and emo-tional relaxation.

Driving motivations, however, differ signif-icantly from country to country, as Fig. 10.2illustrates. To statistically assess these differ-ences, the Scheffe test was employed and theresults are summarized in Table 10.5, showingthat, in terms of adventure and sport, Germantourists are significantly different from eachof the British, the Dutch and the Spanish.

The British consider knowledge as themain driving motivation behind travelling,that is, to discover new places, and socializewith other cultures. The Spanish, on theother hand, tend more toward leisure andrelaxation as the principal motivation.Adventures and sports motivate the Dutch,whereas knowledge and escape/socializationprimarily motivate the Germans.


Table 10.4. Factor analysis results with Varimax rotation of tourists’ motivations for visiting the Algarve.

Factor % of CumulativeComponent loading Communalities Eigenvalue variance %

Factor 1. Adventure and sport 3.34 20.34 20.34To challenge my abilities 1.0720 1.9791To use my imagination 0.9660 2.1670To practise sports 1.0450 1.8394

Factor 2. Knowledge 2.01 12.29 32.63To discover new places and things 0.8180 1.4538To be in a social surrounding 0.9150 1.6033

Factor 3. Escape/socialization 1.86 11.32 43.95To avoid day-by-day stress 0.6050 0.9689To make friends 0.7180 1.4190

Factor 4. Leisure and relaxation 1.86 11.33 55.28To relax physically 1.0510 1.5562To relax mentally 0.6120 1.2987

Source: Correia (2000).


Perception formation

Perception is the mental image that the touristforms of the destination that he or she intendsto visit. In the questionnaire, 20 variables orfeatures were examined addressing organiza-tional, social, psychological and economicaspects of destination perception. Four typesof vacation perceptions were extracted from aPCA, which explained 41.3% of the total vari-ance. Table 10.6 presents the results accordingto the type of vacation or perceptual experi-ence tourists acquire after Varimax rotation.

Table 10.6 shows that consumers carry spe-cific mental images that impinge upon the typeof vacation chosen. Price, social environmentand sport are governing features for sport vaca-tions. Accessibility, cultural exchanges andattractiveness are determining variables for cul-tural vacations, while entertainment andnightlife are important in social-related travel.The impression a tourist develops about aregion’s sun and beaches is linked to climaticand seaside preferences. The various identifiedimpressions of the Algarve were tested accord-ing to country-of-origin. The findings revealsignificant differences between nationalities intheir perceptions of the Algarve. Figure 10.3shows the averages of the four types of vacationpreferred according to each country. Again,the Scheffe test was employed and the resultsare summarized in Table 10.7. There were nosignificant differences between countries withregard to cultural vacation and social vacation. Allfour countries were significantly different from

one another, however, in terms of sports vaca-tion. For the sun and beach, significant differ-ences were observed in all cases other thanbetween Spain and The Netherlands, andbetween the UK and The Netherlands.

It follows from Fig. 10.3 that the Algarve isstrongly associated with being an ideal sportsdestination. This suggests that sun and beachtourism has lost some of its importance,although this appears still to be a strongattractive element in the results for Germany.

Decision phase

Once the tourist has formed his or her mentalimage of a particular destination, the purchas-ing decision is then necessarily dependent onbudget and time constraints; issues that will beexamined in the following sections.

Constraints on tourism demand

Decisive factors affecting tourism demanddepend on income and the amount of avail-able vacation time. In Table 10.8, familyincome as well as allotted vacation time ispresented, according to published statisticsfor 1997.

Table 10.8 shows that Spanish tourists havethe smallest budgets for vacationing purposes.Countries with higher incomes include TheNetherlands and the UK, earning more than80,000 euros. In terms of available time for


–1.0

–0.8

–0.6

–0.4

–0.2

0

0.2

0.4

0.6

0.8

Germany

Spain Netherlands UK

Socialization/escapeLeisure/relaxation Adventure/sportKnowledge

Fig. 10.2. Principal motivation according to country-of-origin.



Table 10.5. Scheffe test for motivations by country-of-origin.

(I) Country of (J) Country of Mean differenceDependent variable residence residence (I�J)a Sig.

Adventure and Germany Netherlands �1.1470560a 0.000sport Spain �0.8619882a 0.000

UK �1.1111642a 0.000Netherlands Germany 1.1470560a 0.000

Spain 0.2850678 0.793UK 0.0358917 1.000

Spain Germany 0.8619882a 0.000Netherlands �0.2850678 0.793UK �0.2491761 0.753

UK Germany 1.1111642a 0.000Netherlands �0.0358917 1.000Spain 0.2491761 0.753

Knowledge Germany Netherlands �0.2122656 0.906Spain �0.0515908 1.000UK 0.0328964 1.000

Netherlands Germany 0.2122656 0.906Spain 0.1606749 0.969UK 0.2451620 0.812

Spain Germany 0.0515908 1.000Netherlands �0.1606749 0.969UK 0.0844871 0.996

UK Germany �0.0328964 1.000Netherlands �0.2451620 0.812Spain �0.0844871 0.996

Escape/socialization Germany Netherlands �0.2736771 0.841Spain �0.0505577 1.000UK �0.1450960 0.975

Netherlands Germany �0.2813991 0.689Spain 0.2736771 0.841UK 0.2231193 0.925

Spain Germany 0.0505577 1.000Netherlands �0.2231193 0.925UK �0.0945383 0.996

UK Germany 0.1450960 0.975Netherlands �0.1285810 0.992Spain 0.0945383 0.996

Leisure and Germany Netherlands 0.2762060 0.846relaxation Spain 0.4354368 0.247

UK 0.3083514 0.603Netherlands Germany �0.2762060 0.846

Spain 0.1592308 0.984UK 0.0321454 1.000

Spain Germany �0.4354368 0.247Netherlands �0.1592308 0.984UK �0.1270854 0.986

UK Germany �0.3083514 0.603Netherlands �0.0321453 1.000Spain 0.1270854 0.986

a The mean difference is significant at the 0.05 level.Source: Correia (2000).


leisure, it is possible to conclude that TheNetherlands possesses the longest vacationperiod, extending up to 32 days, 45.5% morethan the other remaining countries. ForGermany, vacations can vary between 22 and25 days, depending on established contractualemployment negotiations. In the UK, there areno legal limits, although the average vacationperiod is 25 days. Here, years of company ser-vice and seniority can allow for additional vaca-tion time. However, when compared to the restof Europe, the British are given fewer days forpublic holidays. In Spain, the majority get 30vacation days plus weekends.

Choice

Having reached the decision phase, thetourist is faced with a group of interdepen-dent issues. Here decisions such as means oftransportation, lodging and organization ofthe trip need to be made (Table 10.9).

Summer vacation packages dominate pref-erences to the Algarve, although the numberof tourists who choose to travel more indepen-dently has been rapidly increasing. In the UK,intensified competition between travel agen-cies and operators has led to standard discountpolicies. As a result, many consumers makelast-minute decisions to take advantage of spe-


Table 10.6. Factor analysis results with Varimax rotation of tourists’ perceptions of the Algarve.

FactorComponents loading Communalities Eigenvalue % of variance Cumulative %

Factor 1. Sports vacation 2.05 12.08 12.08Safety 0.7110 0.5188Sports 0.5500 0.4780Social environment 0.6110 0.4336Price 0.6050 0.4117

Factor 2. Cultural vacation 1.72 10.09 22.17Attractiveness 0.5350 0.3402Culture 0.5860 0.3533Accessibility 0.6190 0.4964Commerce 0.6240 0.4702

Factor 3. Social vacation 1.66 9.78 31.95Entertainment 0.6570 0.5140Nightlife 0.6570 0.4898

Factor 4. Sun and beach vacation 1.59 9.38 41.33Beach 0.7220 0.5734Climate 0.8020 0.6609


Cultural vacationSports vacation Sun and beachSocial vacation

–1.0

–0.8

–0.6

–0.4

–0.2

0

0.2

0.4


Fig. 10.3. Perceptions according to country-of-origin.



Table 10.7. Scheffe test for perceptions by country-of-origin.

Dependent (I) Country of (J) Country of Mean differencevariable residence residence (I-J)a Sig.

Sport vacation Germany Netherlands �1.2182589a 0.000Spain �0.5288173a 0.001UK �0.4955818a 0.001

Netherlands Germany 1.2182589a 0.000Spain 0.6894417a 0.000UK 0.7226771a 0.000

Spain Germany 0.5288173a 0.001Netherlands �0.6894417a 0.000UK 0.0332354 1.000

UK Germany 0.4955818a 0.001Netherlands �0.7226771a 0.000Spain �0.0332354 1.000

Cultural vacation Germany Netherlands �0.1013417 0.995Spain 0.2141279 0.773UK 0.0230230 1.000

Netherlands Germany 0.1013417 0.995Spain 0.3154696 0.519UK 0.1243647 0.983

Spain Germany �0.2141279 0.773Netherlands �0.3154696 0.519UK �0.1911049 0.813

UK Germany �0.0230230 1.000Netherlands �0.1243647 0.983Spain 0.1911049 0.813

Social vacation Germany Netherlands 0.4428835 0.190Spain 0.4062673 0.141UK 0.2900401 0.466

Netherlands Germany �0.4428835 0.190Spain �0.0366162 1.000UK �0.1528434 0.965

Spain Germany �0.4062673 0.141Netherlands 0.0366162 1.000UK �0.1162272 0.979

UK Germany �0.2900401 0.466Netherlands 0.1528434 0.965Spain 0.1162272 0.979

Sun and beach vacation Germany Netherlands �0.9411064a 0.000Spain �1.1964860a 0.000UK �0.7771671a 0.000

Netherlands Germany 0.9411064a 0.000Spain �0.2553796 0.689UK 0.1639393 0.930

Spain Germany 1.1964860a 0.000Netherlands 0.2553796 0.689UK 0.4193189a 0.035

UK Germany 0.7771671a 0.000Netherlands �0.1639393 0.930Spain �0.4193189a 0.035

* The mean difference is significant at the 0.05 level.Source: Correia (2000).


cial promotional deals. The decision of goingto a travel agency, or not, is influenced by geo-graphical proximity, accessibility and theintended means of transportation. The choiceof destination is often correlated with themode of transportation to be used, the mostfrequent being air travel. The growing use ofair travel is a result of the increased number ofairlines, the decrease in real airfares, as well asgrowing European economic integration.

The Spanish usually rely on the car for theirvacation to the Algarve, travelling in groups ofthree or more people and preferring geo-graphically proximate destinations. In terms oflodging, hotels dominate tourist preferenceswhen travelling to the Algarve, which reflectsan increase in consumer purchasing power.

Post-purchase phase

The degree to which the tourist develops asense of satisfaction, and the probability ofreturning to the destination or of recommend-ing the Algarve to relatives and friends, consti-tutes the principal decision-making behaviourduring the post-purchase phase (Table 10.10).

Because tourists from the UK and Spainreveal a high level of visitor satisfaction,repeated purchases are likely to occur for arelatively higher proportion of these tourists.In contrast, however, The Netherlands revealsa lower level of satisfaction towards theAlgarve, with repeated purchase being uncer-tain for approximately 62% of the touristsbased on the results of the survey.


Table 10.8. Constraints on tourism demand according to country-of-origin.


Earned income ( in percentage terms)Up to 20,000 Euro 9.8 24.5 12.5 16.820,000–80,000 Euro 53.2 47.9 46.9 45.5�80,000 Euro 28.7 9.9 31.9 32.8Unanswered 8.3 17.7 8.7 4.9

Temporal constraints Holidays 22–25 days 24–25 days 32 days 25 daysNational holidays 9 days 14 days 6 days 8 days

Source: DGXXIII (1997, in DGT 1998b), ETM (1998a–d).

Table 10.9. Choice according to country-of-origin.


ChoiceOrganization of trip Travel agency Travel agency Travel agency Travel agency

or self-organizedMeans of transportation Air Car Air AirType of lodging Hotels Hotels, friends Hotels Hotels

and relatives


Table 10.10. Post-purchase evaluation according to country-of-origin.


Level of satisfactionIntention of returning to Probably Definitely Not sure Definitelythe Algarve

Probability of recommending Probably Definitely Definitely Definitelythe Algarve



Germans reveal some reluctance in return-ing to the same destination. This reluctanceseems to derive from the fact that approxi-mately 67% are in doubt as to whether theyshould recommend the Algarve to others.Surprisingly, although the Dutch tend to beonly one-time visitors to the Algarve, theyclaim they would recommend it to others.Their intention of not repeating the visit isnot necessarily evidence of dissatisfaction, butrather a preference to discover new places(particularly since the Dutch are motivated byknowledge). Finally, the survey results indi-cated that the probability of tourists fromSpain and the UK recommending Algarve asan ideal destination was higher than 50%.

Conclusions and Perspectives for FutureResearch

In this paper, the three principal phases ofdecision making were tested with exploratorydata techniques. The methodology appliedallows us to make important inferences aboutthe consumer decision process. It is assumedthat, when the consumer has decided totravel, he or she has acquired adequate infor-mation about the possible combinations ofgoods and services available, and is well awareof the destination options. Assuming thattourists are heterogeneous, the Scheffe testwas used to test significant differences accord-ing to the nationality of origin. The resultsconfirmed the hypothesis that tourist deci-sion making behaviour to the Algarve differssignificantly by country-of-origin.

Bearing in mind that a consumer behav-iour can be structured into three mainphases, pre-decision, decision and post-pur-chase evaluation, the following conclusionsfrom the empirical analysis can be derived:

1. In the pre-decision phase, the followingwere observed:

● the Mediterranean region continues tobe a preferred destination area byEuropean tourists, despite a growingtendency towards new destination expe-riences;

● travel brochures and recommendationsfrom reference groups serve as impor-tant sources of information;

● primary motivations are typically adven-ture/sport, knowledge, escape/social-ization and leisure/relaxation;

● the sun and the beach predominate asthe leading motivation among vacationtourists for the Algarve;

● residential area, social class, age groupand household number constitute deci-sive factors in decision making;

● recommendations from family andfriends as well as previous experiencebear significant influence in decisionmaking;

● PCA analysis indicates that the Algarveis strongly considered as an ideal sportslocation.

2. In the decision phase:● the decision to use travel agencies is

conditioned by geographical proximityand accessibility;

● travel packages are preferred bytourists particularly during the summerseason, although there has been anincrease in independent travel;

● the means of transportation chosen isdependent on the geographical prox-imity of the destination;

● hotels dominate as the preferredchoice for accommodation.

3. In terms of the post-purchase evaluationphase, it was found that the UK and Spainreveal a higher level of satisfaction towardsthe Algarve, whereas Germans reveal somedissatisfaction. The Dutch, being risk-takers(generally searching for new destinations)may not return to the Algarve, but will nothesitate to recommend the Algarve to others.

This study has a number of limitationswhich suggest possible directions for newresearch. First, similar surveys could be carriedout among potential tourists in their countries-of-origin since the present study has surveyedonly visitors (but not non-visitors) to theAlgarve. Second, the questionnaire could havebeen stratified by age, to investigate how age,and related factors such as the family life cyclestage, might affect tourist decision-makinginvolving the Algarve. Lastly, the meaning ofmotivations and perceptions could be differen-tiated according to the respondents’ preferredtourist activities. Additional research is neededto confirm the results of this study, and to clar-ify the above-mentioned issues.



References

Bentler, P. and Speckart, G. (1979) Models of attitude behavior relations. Psychological Review 86, 452–464.Correia, A.J.H. (2000) A procura turística do Algarve. Mimeo, PhD thesis, Algarve University.Crompton, J. (1992) Structure of vacation destination choice sets. Annals of Tourism Research 19, 420–434.Crompton, J. and Ankomah, P.K. (1993) Choice set propositions in destination decisions. Annals of Tourism

Research 20, 461–476.Crouch, G.I. (1994) The study of international tourism demand: a survey of practice. Journal of Travel

Research 32, 41–55.DGT, Direcção Geral de Turismo (1998a) Linhas Orientadoras para a Politica de Turismo em Portugal.

Ministério da Economia, Secretaria de Estado do Turismo, Lisboa.DGT, Direcção Geral de Turismo (1998b) A Sazonalidade do Turismo em Portugal. Ministério da Economia,

Secretaria de Estado do Turismo, Lisboa.ETM, Euromonitor (1998a) The Market for Travel and Tourism – United Kingdom. Report 19.1/Section 4,

Euromonitor, Luxemburg.ETM, Euromonitor (1998b) The Market for Travel and Tourism – Spain. Report 19.1/Section 7, Euromonitor,

Luxemburg.ETM, Euromonitor (1998c) The Market for Travel and Tourism – Germany. Report 19.1/Section 6,

Euromonitor, Luxemburg.ETM, Euromonitor (1998d) The Market for Travel and Tourism – Netherlands. Report 19.1/Section 2,

Euromonitor, Luxemburg.Instituto Nacional de Estatística, 1989 to 1998, Estatísticas do Turismo, Lisboa.Lim, C. (1997) An econometric classification and review of international tourism demand models. Tourism

Economics 3, 69–81.Matos, A.I.M. (2000) A modelização econométrica da procura turística em Portugal. Mimeo, Master’s

degree thesis, University of Oporto.Middleton, V.T.C. (1994) Marketing in Travel and Tourism. Butterworth Heinemann, London.Moutinho, L. (1982) An investigation of tourist behavior in Portugal: a comparative analysis of pre-deci-

sion buying and post-purchasing attitudes of British, American and West German tourists. Doctoralthesis, University of Sheffield.

Ryan, C. (1994) Leisure and tourism: the application of leisure concepts to tourist behaviour – a proposedmodel. In: Seaton, A.V. (ed.) Tourism the State of the Art. John Wiley & Sons, Chichester, pp. 294–307.

Silva, J.A. (1991) O turismo em Portugal. Uma análise de integração micro-macroeconómica. PhD thesis,ISEG-UTL, Lisbon.

Silva, J.A. (1995) A modelização da procura turística em Portugal: um ensaio econométrico. Revista doInstituto de Formação Turística pp. 20–31.

Solomon, M.R. (1996) Consumer Behaviour, 3rd edn. Prentice Hall, Englewood Cliffs, New Jersey.Syriopoulos, T. (1995) A dynamic model of demand for Mediterranean tourism. International Review of

Applied Economics 9, 318–336.Syriopoulos, T. and Sinclair, M.T. (1993) An econometric study of tourism demand: the AIDS model of US

and European tourism in Mediterranean countries. Applied Economics 25, 1541–1552. Um, S. and Crompton, J.L. (1990) Attitude determinants in tourism destination choice. Annals of Tourism

Research 17, 432–448.Woodside, A.G. and Sherrell, D. (1977) Traveler evoked, inept, and inert sets of vacation destinations.

Journal of Travel Research 16, 14–18.World Tourism Organization (1999) Tourism Highlights, 1998. WTO, Madrid.



Chapter elevenThe Consumption of Association Convention Sites:Preliminary Results from a Study of Site Choice

Geoffrey I. Crouch1 and Jordan J. Louviere2

1School of Business, La Trobe University, Melbourne, Victoria 3086, Australia; 2School of Marketing, University of Technology, Sydney, 1-59 Quay Street, Haymarket, NewSouth Wales, Australia

Abstract

The conventions industry has grown to become a very important part of the global tourism and hospitalitysector. For cities, particularly, conventions have become one of their principal target markets. For this rea-son, Convention and Visitor Bureaux typically focus a great deal of their time and effort into wooing largeassociations that are looking for an attractive host convention site. This chapter reports some preliminaryresults from a study which has sought to model the convention site selection choice process. It finds that,although the characteristics of the meeting facilities are particularly important, an attractive host site mustoffer strengths in a broad range of other attributes, if the site is to be successful in an increasingly competi-tive environment.

Introduction

Many Destination Management Organiza-tions (DMOs) in the tourism industry, partic-ularly Convention and Visitor Bureaux(CVBs), today place a great deal of emphasison targeting the meetings and conventionsmarket segment. The task of these organiza-tions is to persuade the corporations or asso-ciations intending to hold meetings orconventions that the destination they repre-sent is the best host site for the event.

Meetings and conventions have grown innumber, size and frequency. The range of

potential host sites has also grown signifi-cantly as destinations have responded bybuilding or expanding convention centresand facilities, as hotels have also added andimproved such facilities, as globalization hastransformed internationally the scope ofmany corporations and associations, and assmaller cities and towns (so-called second-tiersites) have recognized the potential economicbenefits of this market segment and havebegun to compete successfully for a share ofthis activity.

While much is known about the type andrange of factors that influence the choice of a



host site, little is known about the relative sig-nificance of each factor, making it difficultfor destination managers to know where andhow they should invest resources to enhancetheir competitiveness. Specifically, what roledo individual site attributes play in the site-selection decision? How are site attributes‘traded off’ in this process? How does theprocess differ as a function of the corpora-tion’s or association’s characteristics?

This chapter reports preliminary resultsfrom the first stage of a research programmedesigned to answer these questions withrespect to association-organized conventions.The competition that occurs between destina-tions for the meetings and conventions mar-ket segment in the travel and tourismindustry is a microcosm of the issue of desti-nation competitiveness in general. This par-ticular study therefore also draws upon theexperience of Ritchie and Crouch (2003)with respect to their general conceptualmodel of destination competitiveness.

The Meetings and Conventions Industry

Many destinations have increased the empha-sis in their marketing strategies on the meet-ings and conventions industry or travel andtourism market segment. For example, theAustralian Commonwealth Department ofTourism’s National Tourism Strategy (1995)recognized ‘the meetings, incentives, conven-tions and exhibitions (MICE) industry as hav-ing significant growth potential’ (p. 1). Forexample, Australian ‘data indicate that, ingeneral, there has been a rapid increase inthe number of international visitors who havecome to Australia explicitly to attend a con-vention or conference’ (Peters and Jones,1996, p. 4).

Already large, the industry is expected tocontinue to expand. It is easy to understand,therefore, why destinations today covet thismarket segment. Yet, although trade and pro-fessional magazines on the meetings and con-ventions industry frequently publish‘checklists’ of the various factors that profes-sional meetings managers or conferenceorganizers use to assess the suitability andattractiveness of alternative potential host

sites, there has been surprisingly little system-atic, empirical or academic study (Fortin andRitchie, 1977; Var et al., 1985; Fenich, 1992;Witt et al., 1992; Oppermann, 1994; Rockettand Smillie, 1994; Zelinsky, 1994; Clark andMcCleary, 1995). Abbey and Link (1994,p. 283) comment that:

Despite the importance of this market segmentto both individual properties and host cities, lit-tle research has been undertaken on its struc-ture and workings. This lack of information is ahandicap to operating managers and tourismofficials responsible for marketing and promot-ing their products and services … While thislack of research is, on the one hand, a hin-drance to the convention and meeting industry,it presents a promising opportunity forresearchers. Convention and meeting researchis, for the most part, an untapped market forresearchers. Considerable work is needed toincrease understanding of this important seg-ment of the tourism industry.

As noted above, since association-organizedconventions dominate the industry, andbecause corporations differ fundamentallyfrom associations in the way they organizeconventions (Fortin et al., 1976), this presentstudy addresses the issue of convention siteselection by associations. It does not studycorporate conventions or meetings.

Convention Site Selection

Crouch and Ritchie (1998) undertook anextensive review of the literature in order toidentify and evaluate the extent of knowledgeconcerning the factors which are believed toinfluence the choice of convention site byassociations. The research questions thatguided their work were as follows:

● What factors influence the choice of con-vention site?

● How important are each of the site selec-tion factors and how are trade-offs madebetween factors?

● Who participates in the site-selection deci-sion?

● What is the relative influence of each par-ticipant?

● Is the site selection process a function ofcertain association characteristics?

136 G.I. Crouch and J.J. Louviere


● How can destinations enhance their com-petitiveness recognizing the varied controlthey can exert over each principal site-selection factor?

● What are the dynamics of convention siteselection by time and across destinations,associations, and other stakeholders?

On the basis of the 64 studies revealed intheir review, several categories of site-selec-tion factors were identified as illustrated intheir conceptual model of the convention siteselection process (Fig. 11.1). This model illus-trates the influence of these site-selection fac-tors, and consists of five principal steps. Step1 (preplanning) occurs before alternativehost sites are identified and analysed. Itincludes issues such as setting conventionobjectives, formulating a preliminary budget,establishing possible dates for the conven-tion, etc. Alternative sites are then evaluatedin step 2. This might involve actual siteinspections, receiving bids from competinghost sites, liaison with local association chap-ters and CVBs, collecting information abouteach site on key selection criteria such as sizeof meeting facilities, air access, range of

accommodation, the attractiveness of the siteenvironment, etc. Step 2 ends with a recom-mendation from the association’s meetingsmanager or planner, or a committee assignedthe task of investigating alternative sites. Thefinal decision (step 3) from among the alter-natives is usually made separately from thesite analysis and recommendations stage (step2), since this becomes a decision usually forthe board or executive of the associationrather than the meeting planner or site com-mittee. The executive may opt for a site thatwas not the first preference or recommenda-tion arising from step 2. At this stage, politicalissues can often intervene to shape the finaldecision and is best represented as an exam-ple of a business (or organizational) buyingprocess (Kotler et al., 1998).

After the convention is held (step 4),either implicitly or explicitly, the conventionand site is evaluated (step 5) to see whatlessons need to be learnt before the next con-vention is planned.

The intervening factors in this processinclude various antecedent conditions such asassociation, member and executive character-

The Consumption of Association Convention Sites 137

Step 1

Step 2

Step 3

Step 4

Step 5

POST-CONVENTIONEVALUATION

CONVENTIONIS HELD

SITE SELECTIONDECISION

SITE SELECTIONANALYSIS and

RECOMMENDATIONS

PREPLANNINGCONVENTION

COMPETINGSITES

ANTECEDENTCONDITIONS

Nature of the associationMember characteristicsExecutive characteristicsPast experienceAssociation policiesEnvironmental conditionsConvention objectives

SITE

SELECTION

FACTORS

∑ Accessibility∑ Local support∑ Extra-conference Opportunities∑ Accommodation facilities∑ Meeting facilities∑ Information∑ Site environment∑ Other criteria

Fig. 11.1. A general conceptual model of the site-selection process (from Crouch and Ritchie, 1998).


istics, past convention experiences, policies(such as geographic rotation of sites), conven-tion objectives and environmental conditions,such as the prevailing or expected economicclimate or, as we have seen recently, acts of ter-rorism and the SARS virus. Further interven-ing variables include the role and actions ofcompeting sites, and the profile of alternativesites in terms of the key site-selection attrib-utes – the focus of this research.

Research Programme

The previous research discussed above andsummarized in Fig. 11.1 has identified thebroad range of site attributes that the evi-dence suggests may play an important role ininfluencing the selection of a host destinationfor association-organized conventions. At thisstage, however, there is little solid evidenceindicating the relative importance of the vari-ous site-selection factors in general, or in par-ticular circumstances. Hence, the aim of thispresent study is to address this need by model-ling host destination convention site selection.

At this stage of the research programme,this chapter reports some of the preliminaryresults from an Australian study. The approachwill then be adapted for application to an inter-national study of convention site selection. Asthe conventions industry today is lucrative andglobally competitive, the results of this researchshould be of widespread interest.

Research Approach

Choice modelling

The present study adopts this Discrete ChoiceApproach (DCA) for reasons outlined inCrouch and Louviere (2001). Choice model-ling requires choice data that can be obtainedby observing and recording real or actualchoices or choices made in response to hypo-thetical options. Observations of choicesreveal the preferences of those making thechoices; choices collected in real markets typ-ically are known as Revealed Preferences(RP), whereas choices collected in hypotheti-cal markets are known as Stated Preferences(SP) (Louviere et al., 2000).

RP choice data offer the advantage of cer-tainty with regard to actual choice behaviour,but suitable RP data often are unavailable.Moreover, basing choice models solely on RPdata can be disadvantageous due to inade-quate information about choice options con-sidered but rejected, information thatpertains only to preferences for existingoptions, and highly collinear data that makesit difficult to decompose effects or even pre-sents identification challenges. ‘These char-acteristics of markets and the RP dataobservations taken in them unfortunatelysuggest that very rarely will RP data be ofmuch use for modelling purposes’ (Crouchand Louviere, 2001, p. 72).

Although the meetings and conventionsindustry collects some RP data, it generallydoes not provide information on options con-sidered but rejected, rendering it unsuitablefor this type of analysis. Hence, this study setout to gather SP data through the use of asurvey designed as a choice experiment.

Depth interviews

Before preparing the choice instrument, aseries of 25 depth interviews were conductedwith meeting planners who had experienceevaluating and recommending potential hostconvention sites. The aim of these interviewswas to assess specific site attributes for inclu-sion in the choice instrument from amongthe factors listed in Fig. 11.1.

Experimental design

In the experiment, each respondent wasasked to indicate their choice in a series ofchoice tasks in which possible hypotheticalconvention sites were described in terms ofimportant site-selection attributes.

The choice experiment described differ-ent hypothetical sites defined by 20 attributes;each attribute in turn was described by two,four or six levels, resulting in a 210 � 49 � 61

fractional factorial design of 128 treatmentsdivided into eight blocks (survey versions) of16 scenarios.



The present study only considers conven-tion site selection within Australia; hence, cer-tain site attributes that might matter insituations where sites are competing interna-tionally were considered irrelevant forAustralian choices. For example, there is littlevariation in site safety and security inAustralia, but this may well be a key decisionattribute internationally.

Choice survey

The survey consisted of three parts. Thefirst part included a set of questions aboutthe survey participant and the associationwith which the participant was involved inthe role of evaluating host sites in Australiafor the most recent convention, whether asan employee of, or as a consultant to, thatassociation.

The second part of the survey contained16 scenarios, each describing a hypotheticalconvention site. The participant was asked toevaluate the information provided about eachsite and answer two questions: (i) whether theplanner would/would not recommend thatthe board of the association consider that siteoption for the association’s next convention;and (ii) was the site described better than,worse than, or about the same as the last con-vention site chosen. An example of one sur-vey scenario is in the Appendix. This chapterpresents analysis and results for responses tothe first question.

The third part of the survey obtainedinformation about the last convention siteand its attributes, and was structured in sucha way that it closely resembled the format ofpart 2 to facilitate data collection.

Sample

We identified 257 Australian meeting plan-ners with convention site selection experi-ence. Of these, 200 were eligible toparticipate in the choice experiment and 134initially agreed to do so. We ultimatelyreceived 86 completed responses by the cut-off date, resulting in a usable response rate of43% (86/200).

Preliminary analysis

Our preliminary analysis examined the rela-tionship between the subjects’ choices andthe levels of each site that were varied in theexperiment. The first scenario response ques-tion simply asked if the subject would/wouldnot recommend the site for the next conven-tion. We evaluated the relationship betweeneach attribute and this (choice) response byconducting a series of cross-tabulations.

Each attribute appears at two or four dis-crete levels (and one at six discrete levels),and hence a choice experiment can beviewed as a large, incomplete contingencytable. We calculated the Pearson chi-squaredstatistics associated with each cross-tabulationof the choice responses against the levels of aparticular attribute to test the null hypothesisof probabilistic independence. In addition,because we hypothesized that site choicedepends upon the value of each attribute,Goodman and Kruskal tau statistics andUncertainty Coefficient directional measureswere calculated using the choice responsevariable as the dependent variable.

For each statistically significant relation-ship, we examined the direction of the rela-tionship between each attribute and theresponses by plotting the natural logarithm ofthe ratio of choice odds against the attributelevels varied in the choice experiment. Thatis, we calculated the marginal choiceresponses associated with each attribute level,holding everything else constant because theexperimental design is orthogonal. Thisapproach is model-free in the sense that thechoice responses are whatever they are, andhence the relationships that are uncoveredare those that empirically underlie the dataregardless of the process that a researchermight use to represent the systematic andrandom utility components.

For example, if the random component isdistributed independently and identically asan extreme value type I random variate, theresulting model of the process will be abinary logit model. The binary logit is com-pletely defined by the odds of saying ‘yes’ toeach scenario relative to saying ‘no’, and themodel can be linearized by taking the naturallogarithm of the odds ratio, which is linearly



related to the unknown, true utilities of inter-est (Louviere et al., 2000). For this reason, weexamined the relationship of each numericalattribute with the choice response by graph-ing the log of the odds ratio of the marginalprobability of saying ‘yes’ relative to saying‘no’ for each level of each attribute.

The preceding discussion hopefully makesit obvious that the graphical and preliminarystatistical relationships uncovered in thecross-tabular analysis demonstrate that thissimple approach provides virtually all theinformation that is available about the rela-tionships underlying the response data.

The cross-tabulation results are presentedand discussed in the next section.

Results and Discussion

Table 11.1 presents the results from the cross-tabulation analyses for each of the 20 siteattributes shown in the Appendix.

For each attribute found to be statisticallysignificant in Table 11.1, graphs illustratingthe relationship between the log odds calcu-lated from the cross-tab results and the corre-sponding experimental design codes for eachof the four- or two-level attributes, are pre-sented in Figs 11.2–11.13.

Specific results

Proximity of convention participants to con-ference sites was highly significant as shownin Fig. 11.2, which illustrates that the desir-ability of the site declines as the proportionof conference attendees who need to fly fur-ther to get to the conference increases.Interestingly, however, neither unrestrictedeconomy nor discount airfares impacted sitechoices directly.

Accommodation connected to or part ofthe convention facility is also highly desiredby associations, while off-site accommoda-


Table 11.1. Cross-tabulation results – attributes versus choice responsea.

Pearson Uncert. Variable chi-squared Sig.b Tauc Sig.b Coeff.b Sig.b

Participant proximity 22.279*** 0.000 0.016*** 0.000 0.013*** 0.000Unrestricted airfare 1.547 0.672 0.001 0.672 0.001 0.670Best airfare 0.351 0.950 0.000 0.950 0.000 0.950On-site/off-site accomm. 31.202*** 0.000 0.023*** 0.000 0.018*** 0.000Accommodation range 1.816 0.611 0.001 0.612 0.001 0.611Accommodation rates 9.544** 0.023 0.007** 0.023 0.006** 0.020Taxi time 1.331 0.722 0.001 0.722 0.001 0.719Expected weather 1.331 0.722 0.001 0.722 0.001 0.719Cost of venue 34.983*** 0.000 0.025*** 0.000 0.021*** 0.000Food quality 32.658*** 0.000 0.024*** 0.000 0.020*** 0.000Entertainment opps 5.665** 0.022 0.004** 0.017 0.003** 0.017Physical setting 4.128** 0.042 0.003** 0.042 0.002** 0.042Social/cultural setting 2.281 0.131 0.002 0.131 0.001 0.131Chapter assistance 0.054 0.816 0.000 0.816 0.000 0.816Bureau assistance 0.054 0.816 0.000 0.816 0.000 0.816Exhibition space 19.980*** 0.000 0.015*** 0.000 0.012*** 0.000Plenary room 32.405*** 0.000 0.024*** 0.000 0.019*** 0.000Break-out rooms 8.091*** 0.004 0.006*** 0.004 0.005*** 0.004Ball room 1.350 0.245 0.001 0.246 0.001 0.245A/V facilities 2.834* 0.092 0.002* 0.092 0.002* 0.092

aResponse 1 – ‘would’ versus ‘would not’.bSignificance levels for respective statistics in column to the left.cGoodman and Kruskal tau with the response variable as the dependent variable.dUncertainty coefficient with the response variable as the dependent variable.***Significant at 1% level, **significant at 5% level, *significant at 10% level.



0

0.1

0.2

0.3

0.4

0.5

0 1 2 3

Design code

Cro

ssta

b lo

g od

ds

–1.0

–0.5

00 1 2 3

Design code

Cro

ssta

b lo

g od

ds

–1.5

–1.2

–1.0

–0.8

–0.6

–0.4

–0.2

00 1 2 3

Design code

Cro

ssta

b lo

g od

ds

–1.4

–1.2

–1.0

–0.8

–0.6

–0.4

–0.2

00 1 2 3

Design code

Cro

ssta

b lo

g od

ds

Fig. 11.2. Participant proximity.

Fig. 11.3. On-site/off-site accommodation.

Fig. 11.4. Accommodation rates.

Fig. 11.5. Cost of venue.

Consumer Psych - Chap 11 16/12/03 14:25 Page 141


–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

–1.6

–1.4

–1.2

–1.0

–0.8

–0.6

–0.4

–0.2

00 1 2 3

Design code

Cro

ssta

b lo

g od

ds

Fig. 11.6. Food quality.

Fig. 11.7. Entertainment opportunities.

–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

Fig. 11.8. Physical setting.

tion is not desired. The range of availableaccommodation at the site had little impact,and site desirability drops when room ratesreach high levels. Accessibility of the accom-modation site and the local airport appearsto be unimportant.

In terms of the general convention setting,opportunities for entertainment, shopping,sightseeing, recreation and organized tourswere significant, as were physical site attrib-utes and the social and cultural environment.On the other hand, weather expected during


the conference, and assistance from the localassociation chapter or convention and visitorsbureau, were not significant determinants ofchoice as far as the Australian domestic meet-ings and conventions industry is concerned.

Not surprisingly, the cost of the conventionvenue was a major factor, but it is interestingto note (Fig. 11.5) that while site attractiveness

generally declines as cost increases, the lowestcost levels investigated may have signalledpoor and unattractive convention facilities.Alternatively, it may be that associationsexpect a certain minimum level or standard.

In terms of the convention venue andfacilities, the quality of the exhibition space,plenary room, break-out rooms and the per-


–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

–1.2

–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

–1.2

–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

Fig. 11.9. Social/cultural setting.

Fig. 11.10. Exhibition space.

Fig. 11.11. Plenary room.


ceived food quality were important determi-nants of site choice. In contrast, ballroomand dining facilities were less important, aswere availability of required audio-visual facil-ities on-site.

Conclusions and Future Research

Two principal sets of conclusions can bedrawn from these preliminary results: namely,the substantive findings on the factors impact-ing convention site selection within Australia,and the useful role of cross-tabulation as adiagnostic ‘stepping stone’ toward logitchoice modelling of stated choice data.

The preliminary results provide evidencethat 12 of the 20 site attributes investigatedhave a statistically significant effect on sitechoice. These factors were:

● proximity of the site to convention partici-pants;

● percentage of convention attendees ableto be accommodated on-site with the con-vention venue;

● accommodation conference rates;● cost of the venue;● perceived food quality;● opportunities for entertainment, shop-

ping, sightseeing, recreation and orga-nized tours;

● uniqueness of the physical setting;● uniqueness of the social/cultural setting;● quality of the exhibition space;● quality of the plenary room;● quality of the break-out/session rooms;

and● the available range of audio/visual systems

and facilities.


–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

–1.0

–0.8

–0.6

–0.4

–0.2

00 1

Design code

Cro

ssta

b lo

g od

ds

Fig. 11.12. Break-out rooms.

Fig. 11.13. A/V facilities.


Six of these factors pertain to the conventionvenue and facilities while the other six factorsconcern travel distance, cost and site accessibility(one of three), accommodation location and costs(two of four), and the setting, site environmentand local assistance (three of six). Conventiondestinations therefore need to pay primaryattention to ensuring that their meeting andconvention facilities are highly competitivewith respect to their targeted meetings andconvention market segment. In addition,however, other site features also play animportant role, albeit, secondary but no lesssignificant. Hence, a successful conventiondestination must also offer an accessible,enjoyable and interesting setting. A successfulconvention destination must offer a completepackage if it is to perform well in this increas-ingly competitive market.

The study has also shown that, providingthe experimental design of the study is appro-priate, relatively simple analyses of the datacan provide meaningful insights into the roleof individual variables on choice. This also

provides the researcher with a more detailedknowledge and feel for the data beforeembarking on more sophisticated analysis ofthe data.

The future research will adapt the choiceinstrument for the modelling of conventionsite choice on an international scale.

Acknowledgements

The authors gratefully acknowledge thefinancial support for this project from theCooperative Research Centre for SustainableTourism Pty. Ltd, the Association ofAustralian Convention Bureaux Inc., andthe Australian Tourist Commission. The sup-port of the Melbourne Convention andVisitors Bureau, the Meetings IndustryAssociation of Australia, and the SydneyConvention and Visitors Bureau is also grate-fully acknowledged. Jennifer Laing and FayeIhnativ provided valuable assistance withregard to data collection.


References

Abbey, J.R. and Link, C.K. (1994) The convention and meetings sector – its operation and research needs.In: Ritchie, J.R.B. and Goeldner, C.R. (eds) Travel, Tourism and Hospitality Research: A Handbook forManagers and Researchers, 2nd edn. John Wiley & Sons, New York.

Australian Commonwealth Department of Tourism (1995) A National Strategy for the Meetings, Incentives,Conventions and Exhibitions Industry. Canberra, Australia.

Clark, J.D. and McCleary, K.W. (1995) Influencing associations’ site selection process. Cornell Hotel andRestaurant Administration Quarterly 36, 61–68.

Crouch, G.I. and Ritchie, J.R.B. (1998) Convention site selection research: a review, conceptual model,and propositional framework. Journal of Convention and Exhibition Management 1, 49–69.

Crouch, G.I. and Louviere, J.J. (2001) A review of choice modelling research in tourism, hospitality andleisure. In: Mazanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) Consumer Psychology ofTourism, Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK.

Fenich, G.G. (1992) Convention centre development: pros, cons and unanswered questions. InternationalJournal of Hospitality Management 11, 183–196.

Fortin, P.A., Ritchie, J.R.B. and Arsenault, J. (1976) A Study of the Decision Process of North AmericanAssociations Concerning the Choice of a Convention Site. Quebec Planning and Development Council.

Fortin, P.A. and Ritchie, J.R.B. (1977) An empirical study of association decision processes in conventionsite selection. Journal of Travel Research 15, 13–20.

Kotler, P., Armstrong, G., Brown, L. and Adam, S. (1998) Marketing, 4th edn. Prentice Hall Australia Pty.Ltd, Sydney.

Louviere, J.J., Hensher, D.A. and Swait, J.D. (2000) Stated Choice Methods: Analysis and Appliation. CambridgeUniversity Press, Cambridge, UK.

Oppermann, M. (1994) Modelling convention location choice and participation decision making process:a review with emphasis on professional associations. Les Cahiers du Tourisme, Serie C (188). Centre desHautes Etudes Touristiques, Universite de Droit, d’Economie et des Sciences, Aix-en-Provence.


Peters, D. and Jones, B. (1996) Measuring the MICE Industry. Paper presented at the Australian Tourismand Hospitality Research Conference, Coffs Harbour, 6–9 February, 1996, Bureau of TourismResearch.

Ritchie, J.R.B. and Crouch, G.I. (2003) The Competitive Destination: a Sustainable Tourism Perspective. CABInternational, Wallingford, UK.

Rockett, G. and Smillie, G. (1994) Market segments: the European conference and meetings market. EIUTravel and Tourism Analyst 4, 36–50.

Var, T., Cesario, F. and Mauser, G. (1985) Convention tourism modelling. Tourism Management 6, 194–204.Witt, S.F., Dartus, M. and Sykes, A.M. (1992) Forecasting, modelling, and recall bias: modelling conference

tourism. Travel and Tourism Research Association’s 23rd Annual Conference, Minneapolis, 14–17 June,pp. 116–124.

Zelinsky, W. (1994) Conventionland USA: the geography of a latterday phenomenon. Annals of theAssociation of American Geographers 84, 68–86.



Appendix. Convention Site Scenario: Number 1

Features that describe convention sites Details of this site

Travel distance, Proximity of the site to 10% of members do not need to flytravel costs and convention participants 20% of members under 2 h flying time of the sitesite accessibility 30% of members within 2–4 h of siteto members 40% of members over 4 h from site

100%

Unrestricted economy <2 h: $450airfares available 2–4 h: $600

4+ h: $800

Best discount airfares <2 h: $200available 2–4 h: $300

4+ h: $400

Accommodation, Percentage of convention On-site accommodation: 100%location and costs attendees able to be Off-site accommodation: 0%

accommodated on-site versus off-site

Range of available 3 star: 2 hotel(s)accommodation at or within 4 star: 1 hotel(s)15 min of convention facility 5 star: 0 hotel(s)

Conference rates by class 3 star: $90of accommodation (room 4 star: $120only) 5 star: $180

Accommodation location 10-min taxi riderelative to airport

Physical setting, Opportunities for Few opportunitiesclimate, local entertainment, shopping, assistance, and sightseeing, recreation things to do for and organized toursattendees and Unique physical setting Nopartners in the Unique social/cultural Nogeneral vicinity setting

Expected weather/climate Warm and humidduring the conventionExpected level of assistance Satisfactoryfrom local chapterExpected level of assistance Satisfactoryfrom visitors bureau

Convention venue Cost of venue 50% below national averageitself and facilities Exhibition space Marginal

Plenary room MarginalBreak-out/session rooms MarginalBall room/dining venues MarginalRange of A/V systems/ Available completely on-sitepresentation facilitiesPerceived quality of food Below average

I most likely (tick ONLY one) □ Would □ Would not suggest or recommend that the Board of theAssociation seriously consider this convention site for its NEXT convention.

Compared with the LAST convention I helped organize for the Association, I think that THIS poten-tial site would be (tick ONLY one): □ Better than □ Worse than □ About same as the actualsite that hosted this last one.



Chapter twelveContext and Dynamics of Social Interaction and InformationSearch in Decision Making for Discretionary Travel

Tzung-Cheng (T.C.) Huan1 and Jay Beaman2

1Graduate Institute of Management, National Chia-yi University, 151 Lin-Sen East Road,Chia-yi, Taiwan, R.O.C. 600; 2Auctor Consulting Associates, Ltd, 465 Andra Ct, Cheyenne,WY 82009, USA

Abstract

Tourism decision models draw from consumer psychology. These models focus on intrapersonal mentalprocesses, attitudes, values, socio-demographics, some aspects of experience and information use. Thedynamics of information search and social interaction by an individual as part of travel decision makinghave received little attention. Here the dynamic interaction of an individual with the external environmentfor social interaction and information search purposes and an individual’s style of being involved withsome travel decision options are emphasized. This chapter relates decision making to an operator model.Implications of information search and social interaction potentially occurring over time resulting ininterim decisions on what to do are discussed. A final decision is seen as an end that depends on the deci-sion process. The discussion shows decisions being interpreted and modelled as involving a single compen-satory decision when the final decision may depend on interim decisions, some of which may not becompensatory. A simple question–answer approach is used to show the implications of the ideas and theirapplication to survey research.

Introduction

Data that Taiwan collects from internationalvisitors when they depart1 is meant to servemany purposes. Making inferences about thetravel decisions of visitors to Taiwan fromdata that do not include information on howthe person surveyed contributed to thedecision to come can be problematic becausedata relevant to analysis are not available.Thought about data needed resulted in think-

ing about decision making by individuals andhow they were influenced by their socialinteraction and information search.

The tourism literature on informationsearch is extensive (e.g. Fodness and Murray,1999). Regarding social interaction, a classicsocial psychology text, Secord and Backman(1964, p. 1), states that: ‘The most distinctivefeature of human life is its social character.People do things in concert; … And they reactto one another … The behaviour of an individ-



ual in the presence of another person is atonce a response and a stimulus to the otherperson.’ Though Bagozzi (2000) is specificallyconcerned with group decision making, heprovides broad coverage of the general con-sumer behaviour and of other social scienceliterature in making the point that social inter-action does not receive adequate attention inthe consumer decision making literature.

A general flow diagram for the holiday dis-cretionary trip decision process is embeddedin Fig. 12.1, as boxes down the centre of thefigure. Similar process diagrams are availablein the consumer research literature (e.g.Engel et al., 1978, chapters 8–9) but Gitelsonand Crompton (1984) caution againsttourism decisions being treated as routinepurchase decisions. For research, the general

150 T.C. Huan and J. Beaman

Fig. 12.1. Travel decision-making considering social interaction.


process shown has been enhanced in a varietyof ways (e.g. Woodside and Lysonski, 1989;Crompton and Ankomah, 1993; Woodsideand MacDonald, 1994; DeCrop, 1999; Kingand Woodside, 2001; McGuiggan, 2003).Gnoth (1997; see also Harrill and Potts,2002) gives a particularly wide-ranging andinsightful social psychological perspective onmotivation and travel. Action andContemplation appear in the figure, since theextent to which particular activity and desti-nation decisions are made en route or reflectprevious experience is important in under-standing and modelling decisions (Hyde,2003; Kozak et al., 2003).

In the literature on travel decision mak-ing, the selection of options and their consid-eration is recognized as a critical matter (e.g.Crompton and Ankomah, 1993; Woodsideand MacDonald, 1994; DeCrop, 1999). Theconcept of awareness/evoked sets plays animportant role in the discussion of decisions.How these sets arise and how consideration ofthem does, or does not, result in a particularpurchase is an important theme. The litera-ture makes it clear that, as a consequence ofthe urge to travel, a sequence of evoked setsmay be considered (e.g. including early, lateand action). The decisions to change evokedsets and a final decision need not be compen-satory. A decision to modify an evoked setbeing considered can result from findingsalient information or making a pivotal socialcontact. Because getting information andcontacting people occur sequentially in time,one event, e.g. talking with somebody, candirect a decision toward a particular outcomethat is not even an option if the contact is notmade because of information found while try-ing to establish contact.

In Fig. 12.1 there are large arrows withtext in them. Some have the symbol ↔, show-ing two-way communication with the environ-ment. This indicates that information searchand social interaction (e.g. to get a travelcompanion) may occur. However, in thetourism literature, reference to social influ-ences on decision making typically2 is toeffects associated with categorical variablessuch as family status (e.g. Heung et al., 2001,and references therein). Being in a house-hold with four members may mean you are in

an aggregate associated with taking certainkinds of trips. This aggregate result does notelucidate how family members contribute to atravel decision.3

The quotation from Secord and Backmanimplies that recognizing the role of social interac-tion in behaviour is necessary to understanding it.DeCrop (1999, 2001) is a key proponent ofthe need to consider social interaction in tripdecision making. By obtaining longitudinaldata on summer holiday trip decision mak-ing, in which information on social motives,decision participation with others, etc. isavailable, he has bridged a gap recognized byWoodside and MacDonald (1994) andCrompton and Ankomah (1993).Furthermore, he specifically recognizes thatthe social interaction that takes place in rela-tion to a decision to travel depends on theindividual and a context. Types of participa-tion in travel decisions are identified. Somepeople are happy to go along with anotherperson’s suggestion for certain travel deci-sions. When they agree to do what is sug-gested, acceptance is all there is to theirdecision making. For people who do not justaccept an initial option, information search isstill not something everybody does. Somepeople search for information when particu-lar travel is being considered while othersmay not. However, some people are happy tolet other people take the initiative in sortingout travel options. For repeated travel, howmuch searching is usually needed? A conclu-sion that has been reached is that in generalless information may be sought than the liter-ature suggests (DeCrop and Snelders, 2003).Some other researchers have addressed spe-cial implications of interpersonal dynamics(Woodside and Singer, 1994; Thornton et al.,1997; Zalatan, 1998).

Objectives and Research Strategy

Analysis that deals with aggregate effects (e.g.being in a family of five or more members) ismacroanalysis while dealing with an individualand a particular decision making process ismicroanalysis. An objective of this article is tomake specific use of the operator expressionsgiven in Fig. 12.1 (e.g. C(O(pk,t1),T,E,R) •

→

←

Social Interaction and Information Search in Decision Making 151


O(pk,t2) ) in microanalysis of decision makingbehaviour. The discussion of operator expres-sions is used to suggest concepts and to raisequestions about individuals’ social interactionand information search in making a decisionto travel. The series of interim decisions thatcan arise in reaching a final decision is dis-cussed in terms of repeated use of theConsideration operator. There is examinationof the implications of ideas presented forthree general areas of tourism research.

An Operator Model

The operator expressions in Fig. 12.1 providea mathematical/logical symbolism to use inexamining travel decision making. This sym-bolism is used to take a somewhat formalapproach to analysis. Formal systems appear inthe consumer behaviour literature (e.g. Engelet al., 1978, chapter 20) and in the tourism lit-erature (e.g. Kozak et al., 2003). Their role is tofacilitate logical analysis and thus theory devel-opment, modelling and theory testing. Ashere, their use does not necessitate employingany complicated logic or mathematics.

Some definitions and conventions apply tothis paper. Here, process refers to a series ofactions or operations tending toward an end.Actions in a trip decision process can includecommunicating with potential co-travellersabout expectations, costs, travel arrange-ments, etc.; collecting information about tripoptions; and thinking about the consequencesof doing different things. Tending toward anend has two implications. First, a process startsand later finishes. The ‘start’ can influencethe outcome. The weather or one’s spousecan prompt an urge. The spouse prompting itprobably involves a more compelling reasonto pursue a certain course of action. Second,tending implies that what happens during theprocess can influence the outcome, the end.

That travel decision models should havetheir ‘end’ dependent on the process(Woodside and MacDonald, 1994) can bemisunderstood. Dependency occurs when Aor B holiday is taken depending on whetherpx contacts py or pz first when she gets an urgeto get away for the weekend. Implications are:(i) that to some degree, px will do what one of

her friends likes to do; and (ii) that thefriends like different things. On the otherhand, the end does not depend on theprocess just because: (i) outcomes have dif-ferent probabilities based, e.g. on the deci-sion maker’s and destinations’ attributes; (ii)outcomes are computed using different mod-els for different segments (e.g. Stemerding etal., 1996); or (iii) outcomes apply to peoplewho take a particular approach to participat-ing in a decision when planning for certaintravel, e.g. a family holiday (e.g. DeCrop,1999). In (i), (ii) or (iii), if things that hap-pened during the process were introducedinto the model in an appropriate way,4 thenthe outcome would depend on the process.

A simplified picture (see Reed, 1996, foran elaborate one5) of what goes on in themind when considering travel treats mentalfunctions as operations defined by computerprograms. Such computer programs, forexample, control communication to the envi-ronment; recognize, filter, react to and storeincoming communication; control retrieval ofinformation from long-term memory (LTM)for processing; and ‘invoke’ special programsfor problem solving and decision making.Information in LTM can be thought of asdata that is available for retrieval.

Operation, in the context of process, is arrivingat consequences based on antecedents and rulesthat specify what happens in given conditions. Asymbol, sometimes called the operator, is usedto refer to an operation. When the wordsUrge, Formulate, Consideration or Action appearwith a capital and in italics, it is a specific ref-erence to them as operators or to the associ-ated operations. In an operator expressionsuch as U(T,E,R,pk,t0) → s(pk,t0), a specialarrows–dot symbol, •

→

← , or, →, as above, dividesthe expression. In the Urge expression, s(pk,t0)is the consequence or result. It represents astimulus to Formulate travel options. For theexpressions in Fig. 12.1, the initial letter iden-tifies the nature of the operation (e.g. U is forUrge). Inside brackets following this letter onesees operands T, E, R, pk and tx. A particularperson is identified by pk while, regardless ofthe operator it appears in, tx refers to the dateand time since the Urge being consideredoccurred.6 Although the mind does notdivide information related to an Urge by



thoughts (T), input from the external envi-ronment (E), and other stored informationretrieved (R), T, E and R refer to particularinformation in long-term memory, LTM. This‘classification’ of information by ‘origin’ is tofacilitate research conceptualization.

Survey questions about decision makingshould target specific information in LTM.Distinguishing between, e.g., informationthat was obtained from the environment (inE) and related perceptions created by themind using that information (in T) is neces-sary. This is because understanding decisionmaking depends on recognizing how process-ing of environmental information impactsdecisions. Information search (e.g. Fodnessand Murray, 1999, and citations therein) aswell as social interaction can result in ‘raw’information being stored in E (e.g. asepisodic information). Information that wasstored in LTM, e.g. prior to the Urge, and isretrieved to think about travel options relatedto the Urge is in R. Processed information, forexample, resulting from forming an image(e.g. Pike, 2002, and citations therein) istaken to be in T.

When considering operators as computerprograms that do specific things, theoperands are input to that program. Theinputs can be information that is available inLTM, data or parameters that determine howthe program functions. Think of making adecision for a set of options. In Fig. 12.1 thereare ‘decision triangles’ in the boxes forConsideration showing that a decision, ineffect, ends a cycle of Consideration.Therefore, think of a decision operator, Dk,used for selecting one option from several.Each value of k can be for making a choice ina particular way (e.g. one for compensatorydecisions and another for lexicography deci-sions) so based on k the correct Dk is used.

What Fig. 12.1 and the operator expres-sions imply about the dynamics of individualdecision making is straightforward. The largearrows with text in them on the left and rightof the figure, that point towards the boxescontaining expressions, specify the existenceof intrapersonal processes or of communica-tion with the external environment. Thearrows show that an Urge (see box 1 in Fig.12.1) can ‘start’ from a stimulus from the

external environment or an internal one.The notation ‘F(s(pk,t0), T,R)→O(pk,t1)’ showsthat a stimulus leads a person to Formulatetravel options in the form of an initial evokedset O(pk,t1). Formulation of O(pk,t1) is intraper-sonal. Based on → and O(pk,t1) appearing inthe Consideration operator, it moves intoConsideration (box 3).

Consideration has alternative possible out-comes (outputs). This is what the arrows-dotsymbol that divides it into two parts connotes.The possible outputs are: (i) an option beingaccepted for Action as shown by the triple-linearrows to (box 5); (ii) Consideration of revisedoptions (see the triple-line arrows box 3 to 4or 4 back to 4); or (iii) Consideration stopping(see the arrow to ‘Stop’ from boxes 3 and 4).From Fig. 12.1 one sees that messaging out tothe external environment begins inConsideration. Consideration appearing twice inFig. 12.1 (boxes 3 and 4) emphasizes thatmodifying options in an evoked set can beiterative (repeated). This is identified symbol-ically by the Consideration expressions (boxes3 and 4) and by triple-line arrows showingthat a set, O(pk,tn), can be modified toO(pk,tn+1), with O(pk,tn+1) then beingConsidered.

Figure 12.2 provides special insight intoConsideration. In thinking about this figure itis important to recognize why there is con-cern with social interaction and other factorsinfluencing what happens in a givenConsideration step. Certain information (e.g.on rates or availability of reservations) beingobtained or particular social contacts beingmade can prompt evoked set revision. Suchrevision causes an interim decision. One ofthe three possible outcomes of an iteration ofConsideration is selected (stop, accept or con-sider the revised options). An iteration can bethought of as involving getting input, manip-ulating information mentally and making adecision that terminates the cycle. The figuremay seem to suggest that what transpires in acycle (as identified in Fig. 12.2) goes on insome amorphous way but, as introduced ear-lier, what happens during an iteration ofConsideration causes a decision terminating itto be taken.7 That decision need not be compen-satory. It could be lexicographic (e.g. Payne etal., 1990; or Reed, 1996, chapter 14).




Fig.

12.

2.A

Con

side

ratio

ncy

cle:

som

e de

tails

.


In the literature, modifying evoked setsand considering the new set is, in some cases,acknowledged by descriptions of movingfrom an early evoked set to some final set(e.g. Crompton and Ankomah, 1993).Actually, Consideration can end without anyaction other than stopping consideration oftravelling. Of course, people interviewed on atrip moved on to Action on that trip but, asnoted in Fig. 12.1, even deciding to act on atravel option may not result in travel(Woodside and MacDonald, 1994).

Behaviour and the Operator Model

To see the kind of behaviour that is consistentwith the operator expressions for decisionmaking, think about social interaction andinformation search occurring over time inrelation to an Urge. If pk is not travellingalone, communication with potential co-trav-ellers is needed. Any message between peoplemust be considered as having psychosocialgoals, so information transmitted must betaken as directed to achieving certain objec-tives. A message from, e.g., p1, causing an Urgeto travel in p2, can be meant to get afavourable decision on a single option. p1 maytry to limit information search and socialinteraction (e.g. discussion of trip options) byp2 to see that if any options are recognized,only ones favoured by p1 come underConsideration. Clearly, information search andsocial interaction need not result in unbiasedinformation being used by p2. The two peoplemay agree to a strategy for each to pursue forone or several cycles of Consideration. In it, p1could do all the search work because p2 par-ticipates that way with someone like p1 indecisions on the kind of travel being consid-ered (DeCrop, 1999). Assume that in cyclesof Consideration one option is droppedbecause p2 does not like it and another cre-ated to replace it that both parties agree to. Asurvey that does not result in recognizingwhat happened yields an invalid picture ofthe micro-level decision making process.

The example just presented introducessome important matters. A key one is thatpeople may be in different classes/categoriesbased on how they receive and react to a par-

ticular Urge (Woodside and MacDonald, 1994;DeCrop, 1999; McGuiggan, 2003). Then, asintroduced earlier, the outcome of decisionmaking in relation to an Urge can depend oncircumstances. What happens in interactionwith the external environment duringConsideration can cause the process to tendtoward one end or another. What happens is asequence in time in which ‘interim’ decisionsdepend on, e.g. who is around to ‘take a call’.What happens is not just a function ofsources of information used, of number ofsources, or of people contacted. The timingof obtaining particular information or of con-tacts made may be the key to understandingthe flow of the decision process toward anoverall outcome. This is just another way ofstating that travel decision models shouldhave their ‘end’ dependent on the process.

Implications of the Operator Model forTourism Research

Implications of the operator model formula-tion can be divided into three broad cate-gories: estimating models, conceptualizingdecision making as a process and modellingthe process, and survey research involvingunderstanding decision making. Thoughthere are interrelations between these, com-ments are by category.

Estimating models

Modelling decision making using designedexperiments, such as Crouch and Louviere(2001), has not been mentioned earlier in thisarticle. This is because this work and ‘tradi-tional’ analyses using regression techniques,including analysis of variance, are not address-ing the micro-level dynamic matters raisedhere. Furthermore, problems with the kindsof models cited and structural equation mod-els (SEM), sometimes referred to as path orcausal analysis, have been raised regarding: (i)decisions being modelled as compensatory;and (ii) models with inappropriate structuresbeing estimated and accepted (e.g. Bates,1998; Gärling, 1998; Pendyala, 1998). Now, itmay be possible for a choice model (e.g.



Louviere et al., 2000) to be formulated so thatmicro-level dynamics are considered. Other‘non-traditional’ modelling, including the useof neural-network and SEM, offer greatpromise for model development because theyare meant to address dynamic phenomena.This research provides ideas that are usefulfor model development using these methods.

Conceptualizing decision making as a processand modelling the process

Though this chapter is about decision makingas a process, it can only contribute a limitedamount to the large volume of work that needsto be done. It does complement other ongo-ing research that is qualitative (e.g. DeCropand Snelders, 2003) by providing a formalframework built on concepts in the literature.By doing this, other research is facilitated.The authors see survey research that adds toknowledge about attributes related to theoperator model formulation as critical. Thenext section pursues survey-related research.

Moving ahead with formalization of thetravel decision process need not involve usingthe operator model scheme introduced.Applying SEM will foster conceptual develop-ment, because drawing causal path diagramsthat mimic/approximate real-world causalityfacilitates using the method. However, draw-ing realistic diagrams highlights the need forsurvey/observational research that clearlyidentifies what is happening. When oneaccepts the micro-picture of the decisionprocess introduced, then some research needsto be revised to reflect that. For example,Crompton and Ankomah (1993) recognizethe need for data and analysis, such as foundin work initiated in the late 1990s by DeCrop(1999); however, their propositions need to bereformulated. This is because when there arerecognizably different responses to urges thatrelate to types of urge and who is ‘urged’, it isnecessary to assess how the propositions applyin that context. For those people who do notget involved in a particular type of decision,the propositions are virtually irrelevant. Howthey apply to certain other cases will not beknown until more research is completed onindividuals’ participation in types of decisions.

A personal interest of the authors is devel-oping a computerized version of the operatormodel that can be used to simulate behav-iour. Results of survey research are necessaryto create a realistic model. However, withsuch a model, one can create data withknown attributes. These data, for example,can be used to see if ‘surveys’ based on someof the data generated yield valid conclusionswith given analysis procedures.

Survey research involving understandingdecision making

Shoemaker (1994) has selected surveyrespondents only if they played a significantrole in the decision to take a trip. This is notcommon practice. What has been presentedhere actually shows that the nature anddynamics of involvement in the decisionprocess must be understood if individuals’decision making is to be understood. Somepeople have complicated involvement withother people over time in reaching a deci-sion. Asking the person who did not searchfor information about their searching pre-sents the same problems as asking about atti-tudes when attitudes of another personplayed a significant role in deciding what todo. Then, there is the matter of dynamics.How did social interaction and informationsearch contribute to the final decision?Demographic data and numbers or types ofinformation sources do not address theimportance of a particular social contactoccurring or some particular informationbeing obtained at a given time. There possi-bly being a sequence of decisions that aremade in different ways (e.g. compensatory orlexicographic) to reach an end is yet some-thing else that needs to be recognized inanalyses. Many things need to be taken intoaccount to even approximate individual deci-sion making with reasonable accuracy.

It should be clear that a practical way tocollect data that can be analysed to study mat-ters that have been raised is by seeing thatdata collected allow for appropriate segmen-tation. This is achieved by using a priori seg-mentation. Such segmentation refers toplanning data collection, so that responses to



questions asked allow the classification of tripdecision information into the categoriesneeded, so what is happening in the decisionprocess can be understood. Table 12.1 pro-vides a basis for implementing such segmen-tation by presenting potential surveyquestions. One sees that asking what or whocaused an urge to travel matters. This isbecause a stimulus coming from somebody toplease is likely to have a different conse-quence than a stimulus being an advertise-ment on TV. The questions for Formulationmay seem obtuse. However, if a family regu-larly goes somewhere, options may onlybecome an issue with change, e.g. the kidsgrowing up. Furthermore, for an oft-repeatedtrip, limited information search can beexpected. A dimension of this is that notsearching because it is not needed should berecognized as different from letting some-body else search while still depending onsearching being done. The secondFormulation question relates to the matter of avery specific Urge, e.g. going to X for theweekend, not having a specific consequence.One person may acquiesce to a specificrequest to take a certain kind of trip whileanother takes the request as an invitation toraise alternatives. What is in Table 12.1 aboutConsideration should cause one to think about

indicators such as the number of cycles ofConsideration. Of course, information aboutwhat goes on in cycles, particularly whatcauses an interim decision to be taken endinga cycle, is important.

A technical matter is that getting informa-tion about processes can present problemsbecause conventional fixed-length question-naires or questionnaires with fixed substruc-ture for cycles are not really appropriatestructures for data capture. However, movingto new data structures involves learning andthinking through new approaches toanalysis.8 Research is needed on effective datacapture and analysis methodology for micro-level travel decision making data.

An option under study for improving sur-vey research on decision making is formaliz-ing the assumptions associated with Urge,Formulation and Consideration, so that logicaldeduction and simulation can be employed inestablishing viable survey structures. Researchon alternative data structures and on effectivedata capture and analysis methods to use forsuch data is not a priority for the authors untilthere is more progress in codifying what tocollect. For now, taking into account questionslike those in Table 12.1 in planning researchis seen as the viable approach for improvingsurvey data collection.


Table 12.1. Holiday travel decision process: rationale and potential survey questions.

Part of Questions related to Rationale for the information being importantDecision information to collect in some contexts

Urge What or who was the source? Pleasing somebody has special impact.How specific was the ‘Urge ’? Something specific can be acted on directly.

Formulation Is repetition of travel involved? Particular action is to be expected.Was a very specific suggestion Having a specific trip option suggested can just thought provoking? cause a range of thought rather than action.

Consideration Was there really anything to consider? An offer to go with a friend or on a companymay not involve options other than go or not.

What happened during Consideration Modifications to evoked sets can arise from in terms of social interaction, information information search and social interaction. In search to change destinations considered? terms of DeCrop’s (1999) classes, of actual What was their sequence? What pressures information search and of social interaction, were brought to bear or did you feel from it is reasonable to think that for many studies sales people or friends, etc.? the data collected from many respondents

will have little to do with their travel decision(e.g. to be where interviewed).

Were ‘dummy’ options introduced as part If options listed in data are for bargaining, of bargaining? treating them as ‘real’ is not appropriate.


Reflections/Conclusion

This chapter has used an operator model for-mulation in pursuing matters that are as impor-tant in understanding decision making ascollecting data from people who were activelyinvolved in decision making. The discussion ofthe implications of this research for researchon decision making has resulted in recogniz-ing that a critical matter is developing anunderstanding of processes at the micro-level.Progress in understanding individuals’ deci-sion making depends both on getting data andon concept development. Concept develop-ment is necessary so that the data collected areactually useful for elucidating important struc-tural matters. It is reasonable to think thatprogress in understanding structure will resultin capabilities to develop viable process modelsthat can be used in research on survey andanalysis methodology. Therefore, it is hopedthat what has been presented encourages datacollection and model development that servesto improve the understanding of individuals’travel decision making.

Notes

1See, e.g. Huan et al. (2002) for some informationabout Taiwan’s survey of outbound internationalvisitors.

2 ‘Typically’ is used in this paragraph based onthe scarcity of articles actually addressing socialinteraction.3 For example, one has statements of the form:being an X is associated with doing Y more thanmost people. This is a statement about aggregateinfluence not individual behaviour.4 One way to incorporate such influences is todefine causal elements of the process using struc-tural equation modelling (SEM). Whether viableforms of such models can be defined for theprocess is research that can be pursued. 5 For detail one is obliged to read much of thematerial in a book such as Reed (1996). 6 To avoid making a notation that some may findconfusing even more complex, the expressionsapply to a particular urge. An operand couldidentify a particular urge. Then one could havemultiple sets of expressions for a person. Howoperations for different urges might be interre-lated (e.g. because of budget, time conflict, etc.)becomes an issue. The notation is actually simpli-fied in other ways, e.g. by not having subscriptsfor T, E and R showing that it is what is in them ata given time that matters.7 Note that much of the literature assumes thatthere is no iteration so there is input as per Fig.12.2 followed by a decision. Modelling of thedecision, based on the regression methods oftenused, takes the decision to be compensatory.8 A ‘natural’ capture involves data capture as occursfor an object in motion with various things influ-encing its motion. What happens when and howinfluences operate must be recorded to understandwhat is happening to it.


References

Bagozzi, R.P. (2000) On the concept of intentional social action in consumer behavior. Journal of ConsumerBehavior 27, 388–396.

Bates, J. (1998) Reflections on stated preferences: theory and practice. In: Ortúzar, J., Hensher, D. andJara-Díaz, S. (eds) Travel Behaviour Research: Updating the State of Play. Elsevier Science Ltd, Oxford,pp. 89–103.

Crompton, J.L. and Ankomah, P.K. (1993) Choice set propositions in destination decisions. Annals ofTourism Research 20, 461–476.

Crouch, G.I. and Louviere, J. (2001) A review of choice modelling research in tourism, hospitality andleisure. In: Mazanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) Consumer Psychology ofTourism, Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK, pp. 67–86.

DeCrop, A. (1999) Tourists’ decision-making and behaviour processes. In: Pizam, A. and Mansfeld, Y. (eds)Consumer Behaviour in Travel and Tourism. Haworth Press, New York, pp. 103–133.

DeCrop, A. (2001) The antecedents and consequences of vacationers’ dis/satisfaction: tales from the field.In: Mazanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) Consumer Psychology ofTourism, Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK, pp. 333–347.

DeCrop, A. and Snelders, D. (2003) Vacation decision making: an adaptable and opportunistic process(under review).

Engel, J.F., Blackwell, R.D. and Kollat, D.T. (1978) Consumer Behaviour, 3rd edn. Dryden Press, Hinsdale,Illinois.


Fodness, D. and Murray, B. (1999) A model of tourist information search behaviour. Journal of TravelResearch 37, 220–230.

Gärling, T. (1998) Behavioural assumptions overlooked in travel choice modelling. In: Ortúzar, J.,Hensher, D. and Jara-Díaz, S. (eds) Travel Behaviour Research: Updating the State of Play. Elsevier ScienceLtd, Oxford, pp. 3–18.

Gitelson, R.J. and Crompton, J.L. (1984) Insights into the repeat vacation phenomenon. Annals of TourismResearch 11, 199–217.

Gnoth, J. (1997) Tourism motivation and expectation formation. Annals of Tourism Research 24, 283–304.Harrill, R. and Potts, T.D. (2002) Social psychological theories of tourist motivation: exploration, debate

and transition. Tourism Analysis 7, 105–114.Heung, V.C.S., Qu, H. and Chu, R. (2001) The relationship between vacation factors and socio-demo-

graphic and travelling characteristics: the case of Japanese leisure travellers. Tourism Management 22,259–269.

Huan, T.C., Beaman, J.G. and Shelby, L. (2002) Using action-grids in tourism management. TourismManagement 23, 255–264.

Hyde, K.F. (2003) Vacation decision-making: an alternative framework. Paper presented at the ThirdSymposium on the Consumer Psychology of Tourism, Hospitality and Leisure, Melbourne, La TrobeUniversity, January 2003. Also in Tourism Analysis (in review) in a summary version and in this book.

King, R.L. and Woodside, A.G. (2001) Qualitative comparative analysis of travel and tourism purchase-con-sumption systems. In: Mazanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) ConsumerPsychology of Tourism, Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK, pp. 87–105.

Kozak, M., Huan, T.C. and Beaman, J.G. (2003) A systematic approach to non-repeat & repeat tourismwith measurement and destination loyalty concept implications. Journal of Travel and TourismMarketing 12, 19–38.

Louviere, J., Hensher, D. and Swait, J. (2000) Stated Choice Methods. Cambridge University Press, Cambridge.McGuiggan, R.L. (2003) A model of vacation choice: an integration of personality and vacation choices

with leisure constraints theory. Paper presented at the Third Symposium on the ConsumerPsychology of Tourism, Hospitality and Leisure, Melbourne, La Trobe University, January 2003. Alsoin Tourism Analysis (in review) in a summary version and in this book.

Payne, J.W., Bettman, J.R. and Johnson, E.J. (1990) The adaptive decision maker: effect and accuracy inchoice? In: Goldstein, W.M. and Hogarth, R.M. (eds) Research on Judgment and Decision Making.Cambridge University Press, Cambridge, pp. 181–204.

Pendyala, R.M. (1998) Causal analysis in travel behaviour research: a cautionary note. In: Ortúzar, J.,Hensher, D. and Jara-Díaz, S. (eds) Travel Behaviour Research: Updating the State of Play. Elsevier ScienceLtd, Oxford, pp. 35–48.

Pike, S. (2002) Destination image analysis – a review of 142 papers from 1973 to 2000. Tourism Management23, 541–549.

Reed, S.K. (1996) Cognition: Theory and Applications, 4th edn. Brooks/Cole, New York. Secord, P.F. and Backman, C.W. (1964) Social Psychology. McGraw-Hill, New York.Shoemaker, S. (1994) Segmenting the U.S. travel market according to benefits realized. Journal of Travel

Research 32, 8–21.Stemerding, M.P., Oppewal, H., Beckers, T.A.M. and Timmermans, H. (1996) Leisure market segmenta-

tion: an integrated preferences/constraints-based approach. Journal of Travel and Tourism Marketing 5,161–186.

Thornton, P.R., Shaw, G. and Williams, A.M. (1997) Tourist group holiday decision-making and behaviour:the influence of children. Tourism Management 18, 287–297.

Woodside, A.G. and Lysonski, S. (1989) A general model of traveller destination choice. Journal of TravelResearch 27, 8–14.

Woodside, A.G. and MacDonald, R. (1994) General system framework of customer choice processes oftourism services. In: Gasser, R.V. and Weiermair, K. (eds) Spoilt for Choice. Decision-making Processes andPreference Change of Tourists: Intertemporal and Intercountry Perspectives. Kulturverlag, Thaur, pp. 30–59.

Woodside, A.G. and Singer, A.E. (1994) Social interaction effects in the framing of buying decisions.Psychology and Marketing 11, 27–35.

Zalatan, A. (1998) Wives’ involvement in tourism decision processes. Annals of Tourism Research 25,890–903.



Chapter thirteenA Duality in Vacation Decision Making

Kenneth F. HydeManukau Institute of Technology, Private Bag 94-006, Auckland, New Zealand

Abstract

This chapter examines consumer decision making for choice of elements in the independent vacation.Both destination marketing organizations and tourism businesses would benefit from an understanding ofwhich vacation elements are typically chosen by the traveller pre-trip versus which vacation elements aretypically chosen while on vacation. Previous studies on the subject have produced conflicting results. Thechapter seeks to explain these discrepancies by referring to the concept of a traveller’s familiarity with thedestination. Furthermore, the chapter suggests a duality might exist in vacation decision making. Pre-tripdecision making might best be described as deliberate, purposeful and reasoned. On-vacation decisionmaking might at times be described as light-hearted, free-spirited and hedonistic. The extent to whichsuch a duality exists has yet to be established empirically.

Introduction

An understanding of vacation decision makingis at the core of a consumer psychology oftourism. The following is a conceptual studythat examines the sequence in which elementsof the independent vacation are chosen.

The chapter commences with a review ofseveral established models of vacation deci-sion making, and draws conclusions regard-ing the ideal content of a model of vacationdecision making. Baseline statements aremade regarding pre-trip decision making andon-vacation decision making. Discrepanciesare noted in the findings of three previousstudies on the subject – the study presentedby Hyde (2000) at the First Symposium on

the Consumer Psychology of Tourism,Hospitality and Leisure, the study byFesenmaier and Jeng (2000) and the studypresented by King and Woodside (2001) atthe Second Symposium. To explain these dis-crepancies, the concept of familiarity with thedestination is employed. The chapter suggeststhat pre-trip decision making and on-vacationdecision making represent a duality in vaca-tion decision making. The characteristics ofthese two modes are outlined.

Models of Vacation Decision Making

The field of consumer psychology is foundedon the perspective of the consumer as a prob-



lem solver. The Engel et al. (1968) model sug-gests that consumer behaviour consists of fourcomponents: inputs to the decision process,information processing, the decision processand other variables influencing the decisionprocess. The decision process is said to consistof five steps, namely problem recognition,information search, evaluation of alternatives,purchase and post-purchase outcomes.

An alternative perspective is the view of theconsumer as a pleasure seeker. Holbrook andHirschman (1982) pointed out that the seekingof emotional arousal is a major motivation forthe consumption of certain product classes,including the arts, entertainment, sports andleisure pursuits such as tourism. The traditionalview of the consumer as a problem solver mightbe inappropriate to describe these purchases,which are based on satisfying emotional wantsrather than utilitarian needs.

During the 1980s, a number of models ofconsumer behaviour in tourism were pre-sented. The models of van Raaij and Francken(1984), Moutinho (1987), Woodside andLysonski (1989), and Um and Crompton(1990) in particular attracted attention. Eachof these models drew on the perspective ofthe consumer as a problem solver.

van Raaij and Francken (1984) proposed amodel of vacation decision making, repre-sented as a sequence of sub-decisions. Thatsequence is: (i) the generic decision to spendon a vacation; (ii) information acquisition;(iii) joint decision making by husband andwife; (iv) experience of the vacation; and (v)subsequent levels of satisfaction.

According to Moutinho (1987), tourismmarketing is based on understanding how con-sumers perceive destination areas, how theylearn about their tourism options, how theymake travel decisions and how personalityinfluences these decisions. Marketing stimuli,social factors, characteristics of the destinationand other external factors are listed as influ-ences on the choice of travel destination.

Woodside and Lysonski (1989) presented amodel on tourist destination awareness andchoice. The model proposes a sequence ofevents in destination choice, namely aware-ness, affective associations, preferences, inten-tions and choice. Marketing activities, personaland social factors are listed as influences.

According to Um and Crompton (1990),passive information capturing may lead toformation of an awareness set of travel desti-nations. Active search for information on des-tinations, and consideration of suchconstraints as time and cost, will reduce thisawareness set to a smaller consideration set.Attitudes may be formed towards each desti-nation in the consideration set.

Each of these models, however, is limited,in that it seeks to explain a singular decisionregarding the vacation, namely choice of pri-mary destination (Decrop, 1999), but vaca-tions are much more than travel to adestination. Vacation decision making is aparticularly complex and multi-faceted affair,involving a great number of decisions madeover a period of days, weeks or months.

Many independent vacations are multi-destinational, or touring, in nature (Lue etal., 1993). A model of vacation decision mak-ing should consider not only the traveller’schoice of primary destination, but choice ofsecondary destinations as well.

A model of vacation decision makingshould also consider the traveller’s choice ofaccommodation and transport options, choiceof travel route, choice of attractions and activi-ties, and choice of dining and shoppingoptions. Woodside and MacDonald (1994)provided a framework that comprehensivelycovers consumer choice of all these major ele-ments in the vacation. Their model proposesthat choices of tourism-related services areeach interdependent to some degree, but theseresearchers pointed out that there is a need toresearch the sequence of decision makingamongst the elements of the vacation. Doeschoice of secondary destinations precedechoice of attractions and activities, for exam-ple, or does choice of attractions and activitiesprecede choice of secondary destinations?

Baseline Statements

Several initial statements can be made regard-ing the sequence in which elements of theindependent vacation are chosen.

Regarding the decision making processesof the traveller before they depart on vaca-tion, that is their pre-trip decision making, itcan be said that:

162 K.F. Hyde


● all travellers choose their primary destina-tion and means of transport to the destina-tion before departure;

● some travellers undertake deliberate andpurposeful research of their destinationand the vacation elements available;

● some travellers choose some vacation ele-ments – secondary destinations, accommo-dation and transport options, travel route,attractions and activities, dining and shop-ping options;

● some travellers neither research nor decideupon vacation elements before departure.

Regarding the decision making processes thatoccur while on vacation, it can be said that:

● a proportion of all the traveller’s vacationelements will be chosen while on vacation;and

● that proportion is greater than zero;● all travellers research and explore their

destination while on vacation.

It is of considerable interest to business and todestination marketing organizations to knowwhich vacation elements are typically chosenprior to departure versus which elements aretypically chosen while on vacation. If vacationelements are chosen pre-trip, then marketingcommunications efforts and expenditureshould be concentrated on the prospectivetraveller resident at their home. If vacation ele-ments are chosen on vacation, then marketingcommunications efforts and expendituresshould be concentrated on the traveller in situ.

Existing research presents conflicting evi-dence on the sequence in which elements ofthe vacation are chosen.

Hyde’s Model of the Vacation as‘Evolving Itinerary’

According to Hyde (2000), the independentvacation is like experiencing the ‘fun of thefairground’, a freewheeling experience of trav-elling from place to place, relatively unaware ofwhat each secondary destination has to offer.An integral feature of independent travel is theenjoyment the consumer experiences from notplanning the details of the vacation but ratherfrom experiencing the unknown and unex-pected. Rather than following a fixed itinerary,

the elements of an independent vacationevolve as the vacation proceeds.

Hyde studied the vacations of 20 travelparties who were visiting New Zealand for thefirst time. In New Zealand, the primary attrac-tions are geographically dispersed and thetourist must choose a route connecting theirselection of secondary destinations. NewZealand has one of the highest lengths of stayand intra-national dispersion of tourists ofany international destination (Oppermann,1994). Travel parties were interviewed both atthe beginning and at the end of their vaca-tions, to compare vacation research and planswith actual vacation behaviours undertaken.Lengths of vacation varied from 10 to 88 days.

According to Hyde, a plan to visit a seriesof secondary destinations is formulated priorto the vacation commencing. This plan willbe acted upon, but other vacation elementswill not be chosen until the tourist is on vaca-tion. The tourist’s choice of attractions andactivities is generally delayed until the day ofarrival at a secondary destination. The touristtakes advantage of serendipitous opportuni-ties to experience a number of secondary des-tinations, attractions and activities that theyhad neither actively researched nor planned.According to Hyde, the sequence of vacationdecision making is firstly choice of secondarydestinations, then choice of travel route, thenchoice of attractions and activities.

Fesenmaier and Jeng (2000)

Fesenmaier and Jeng (2000) asked partici-pants in an experiment to imagine a 2–4 daydomestic vacation trip, and consider which of14 vacation elements they might plan beforedeparture. Three levels of vacation decisionmaking were identified. A set of core deci-sions were made early in vacation planning;these decisions included choice of primarydestination, location of overnight stay/s andtravel route. A second set of decisions, madeprior to departure but somewhat flexible,included choice of secondary destinations,attractions and activities for the vacation. Ingeneral, the only elements of the vacationdecided upon while on vacation were matterssuch as the choice of where to shop and

A Duality in Vacation Descision Making 163


where to dine. Thus, for a hypothetical trip towhat is probably a familiar domestic destina-tion, most elements of the vacation appearedto have been chosen pre-trip.

King and Woodside’s Model of theVacation as Purchase–Consumption

System

King and Woodside (2001) employed theconcept of a purchase–consumption system(PCS) to model vacation decision making. APCS is a sequence of purchases a consumerundertakes, in which the purchase of oneitem triggers the purchase of several others.This concept provides a framework for under-standing the interrelated nature of decisionstaken in selection of vacation elements.

The King and Woodside study involvedanalysis of the vacations of 68 travel parties tothe Big Island of Hawaii. The Big Island is adestination dominated by a single attraction,the Hawaii Volcano National Park, and thetourists’ vacations varied from 1 to 12 days induration. Data on the travel parties weregathered retrospectively, at the end of theirvacations. The data confirms that: some vaca-tion elements were not planned before thestart of the vacation; some vacation elementsmay be chosen either pre-trip or on-vacation;and some decisions within a tourist’s PCS aredependent on prior purchases that triggerthese later purchases.

Within the vacations of these tourists, Kingand Woodside recognized three levels of deci-sion making. Level 1 decisions were made inthe early stages of vacation planning, andinvolved selection of the primary destination,plus the attractions and activities to visit.Level 2 decisions were also made before thevacation commenced, and involved choice oftravel route and accommodation. Level 3decisions were made while on vacation, andprincipally involved choice of where to dineand where to shop.

A Comparison of Models

Woodside and MacDonald had challengedresearchers to discover the sequence of deci-

sion making amongst the elements of thevacation. It is on this point that the datasets ofHyde, Fesenmaier and Jeng, and King andWoodside differ markedly. Amongst thetourists to New Zealand studied by Hyde,choice of secondary destinations appeared toprecede choice of attractions and activities;choice of attractions and activities were largelymade while on vacation. Amongst the subjectsin Fesenmaier and Jeng’s experiment, choiceof secondary destinations, attractions andactivities were generally made prior to depar-ture. Amongst the tourists to the Big Islandstudied by King and Woodside, choice ofattractions and activities preceded choice ofsecondary destinations; choice of attractionsand activities were largely made concurrentwith the choice of primary destination.

There is, therefore, conflicting evidencefrom existing published studies into thesequence and timing of choice of vacationelements. Potential explanations for thesefindings could include differences in charac-teristics of the destinations visited, differencesin the ease with which vacation elements canbe purchased prior to departure, or differ-ences in the characteristics of travellers tothose destinations.

Familiarity with the Destination

One explanation for these findings thatdeserves further attention is the degree offamiliarity the traveller has with the destina-tion. Familiarity with the destination could bebased on prior visits, or based on considerablepre-vacation information search by the trav-eller. Past research has suggested the impor-tance of familiarity with the destination.

Cooper (1981) observed the vacations ofvisitors to the Channel Island of Jersey. Mostvisitors visited the familiar, top-of-mind attrac-tions in their first 2 or 3 days of the vacation.It was only in the last days of their vacationthat visitors chose to view the island’s lessfamiliar attractions.

In their study of visitors to AlachuaCounty, Florida, Crotts and Reid (1993)noted that many domestic visitors, familiarwith the destination, had chosen their vaca-tion activities pre-trip, while many long-haul

164 K.F. Hyde


international visitors, less familiar with thedestination, chose their vacation activitieswhile on vacation.

Oppermann (1997) examined the vaca-tions of first-time versus repeat visitors toNew Zealand. His data showed that first-timevisitors visited more secondary destinations,attractions and activities than did repeat visi-tors. One obvious explanation for this find-ing is that first-time visitors displayexploratory and variety-seeking behaviour;they seek to experience and understand asmuch of the destination as time, effort andexpense allow. They seek to become familiarwith the destination. Repeat visitors, on theother hand, are already familiar with the des-tination, and may seek to repeat just thosevacation elements they enjoyed on the firstvisit to the destination.

Returning to the datasets of Hyde,Fesenmaier and Jeng, and King andWoodside, familiarity with the destinationmay offer an explanation for these conflictingfindings. The most likely reason that respon-dents in Hyde’s study had planned very fewvacation elements was that they were visitingan unfamiliar international destination. Thevacation to New Zealand was going to be arelatively long one, and as the vacation pro-ceeded, the tourist was exposed to many sec-ondary destinations, attractions and activitiesthey had not previously been aware of.

The most likely reason that the respon-dents in Fesenmaier and Jeng’s study hadplanned most of their vacation elements wasthat they were visiting a familiar domesticdestination.

The most likely reason that respondentsin King and Woodside’s study had pre-planned their choice of attractions, activitiesand travel route was that they were familiarwith the vacation elements on offer in theBig Island before their vacation com-menced. For many of these tourists, theirvacation was dominated by the ‘icon’ attrac-tion of the Hawaii Volcano National Park.Their choices of secondary destinations andattractions to experience were predeter-mined by their desire to visit the VolcanoNational Park. The destination visited wasgeographically smaller; the vacation experi-enced was noticeably briefer in duration. As

a result, the proportion of vacation elementschosen under conditions of unfamiliaritywas smaller than for the tourists on a NewZealand vacation.

This distinction then, between vacationsexperienced under conditions of familiarity,and vacations experienced under conditionsof unfamiliarity, provides a valuable frame-work from which to view vacation decisionmaking. While use of this concept of familiar-ity with the destination may appear self-evi-dent (i.e. ‘the traveller who plans most is thetraveller who knows most’), it may prove avaluable tool in understanding variations inthe sequence in which elements of a vacationare chosen.

Consumers at Play

To date, models of vacation decision makinghave taken the perspective of the consumeras a problem solver (Engel et al., 1968), butthe perspective of the consumer as a plea-sure seeker may also prove valuable(Holbrook and Hirschmann, 1982). Theview of the consumer as a pleasure seekerargues that many consumer choices resultfrom the human need to experience feelingsand emotions. Decision making processesfor choice of leisure and tourism servicesmay revolve around the consumer’s goal ofprovoking feelings and emotions, ratherthan seeking to solve problems. The evi-dence here suggests that consumers at playdisplay pleasure-seeking and variety-seekingbehaviours, and make choices amongst alter-native activities in a relatively light-heartedand unconsidered manner.

Elliott and Hamilton (1991) studied con-sumer choice of evening, out-of-homeleisure activities. The researchers inter-viewed 560 young consumers immediatelybefore they engaged in an evening leisureactivity. Their study suggested that choice ofsuch activities usually involves the use of sim-ple heuristics, such as ‘doing something dif-ferent for a change’, rather than extensivedecision making.

From their intensive study of consumers ofriver rafting, Arnould and Price (1993) pro-vide some intriguing conclusions regarding



decision processes for choice of leisure activi-ties. Consumers of leisure activities may seekintense emotional outcomes but not knowwhich alternatives will produce them. In theirstudy, Arnould and Price found informationsearch for choice of rafting trip to be mini-mal. Even though river rafting costs a lot oftime and money, consumers did not appearto think carefully about their choices.Arnould and Price suggested that consumers’limited expectations may even have con-tributed to the satisfaction experienced, inthe sense that experiencing the unknown wasa source of thrill.

Kemperman et al. (1997), in their study ofa consumer’s choice of activities in a themepark, found strong evidence of variety seek-ing amongst consumers at play. Much likethe diner at a buffet meal, the visitor to atheme park wants to sample a little of every-thing on offer; they want to feel that theyhave not missed out on some important ele-ment the theme park has to offer. The find-ings of Oppermann (1997) regardingfirst-time versus repeat visitors to an interna-tional destination, suggest that travellers toan unfamiliar destination might also displayvariety-seeking behaviour.

There is some evidence then, from studiesof consumers at play, that on-vacation deci-sion making might be characterized by a lessdeliberate, most simplistic decision process,and a desire for variety seeking when choos-ing vacation elements.

A Suggested Duality in VacationDecision-Making

It may well be that the typical processes ofvacation decision making which occur pre-tripare distinct from the decision makingprocesses which occur on vacation. There mayexist a duality in vacation decision making.

Pre-trip vacation decision making for theindependent traveller might best bedescribed as deliberate, purposeful and rea-soned. Many vacationers are highly involvedin the selection of the vacation destination/sand the planning of a vacation itinerary.During this phase, many tourists undertakeextensive information search, including use

of travel books and travel brochures, use oftravel industry personnel and, increasingly,use of the Internet. Information search andcomparison of alternatives may proceed overa period of days, weeks or months. A pri-mary destination is chosen. Secondary desti-nations, travel routes, attractions andactivities may also be chosen. Where thetravel party consists of more than one per-son, negotiation and compromise occurbetween the parties. The perspective of theconsumer as a problem solver is an appropri-ate framework for understanding pre-tripvacation decision making.

On-vacation decision making might at timesbe described as light-hearted, free-spirited andhedonistic. Vacation time represents distinctoccasions in a consumer’s life; the consumeron vacation is a consumer at play (Mergen,1986; Ryan, 2002). Just as a consumer’s choiceof leisure activities can be based on minimalinformation search and use of simple choiceheuristics, so might a consumer’s choice ofoptions while on vacation. Consumers on vaca-tion seek new and novel experiences. They areexplorers. Like visitors to a theme park, con-sumers on vacation can exhibit variety-seekingbehaviours. On vacation, consumers can alsobe more free-spending, and purchase moreitems of a purely hedonistic nature. The per-spective of the consumer as a pleasure seeker isan appropriate framework for understandingon-vacation decision making.

Discussion

This chapter has presented an alternativeframework for understanding vacation deci-sion making. This framework concerns itselfwith the sequence of decision makingamongst vacation elements. It recognizes thatthe sequence of decision making displayedmay vary according to the tourist’s degree offamiliarity with the destination.

The managerial implications of this frame-work are obvious. If we wish to influence aconsumer’s decision making for choice of aparticular vacation element, such as choice ofattraction or activity, then we need to knowwhere in the chronology of the vacation thatpoint of decision making is.

166 K.F. Hyde


As each month goes by, the Internetexpands in volume. Tourists to an unfamiliardestination now have ready access to minu-tiae of detail on the primary and secondarydestinations they may visit. More and moretourists are using the Internet to assist theirpre-vacation planning (Bonn et al., 2001).This raises the possibility that the bulk oftourists will become more familiar with theirintended destinations than was formerly thecase. It raises the possibility that a greater pro-portion of tourists will demonstrate vacationdecision making under conditions of familiar-ity, and a smaller proportion of tourists willdemonstrate vacation decision making underconditions of unfamiliarity. This is a pointworthy of research.

The suggestions made in this chapter areat best tentative. The chapter suggests thatfuture research into vacation decision makingshould recognize a duality in vacation deci-

sion making processes – pre-trip versus on-vacation decision making. The extent of thatduality has yet to be established empirically.In particular, in situ research of the decisionprocesses of consumers on vacation is a valu-able field of investigation.

Both macro-models and micro-models ofvacation decision making are required.Macro-models of the vacation are required,models that can provide a comprehensiveframework encompassing choice of all ele-ments of the vacation. One such model is theGeneral Theory of Tourism ConsumptionSystems (Woodside and Dubelaar, 2002). Amicro-model of vacation decision making isalso required, which describes the precise insitu decision processes employed by the con-sumer while on vacation, for choice of anattraction, activity, retail store, restaurant,accommodation house, secondary destina-tion or travel route.


References

Arnould, E.J. and Price, L.L. (1993) River magic: extraordinary experience and the extended serviceencounter. Journal of Consumer Research 20, 24–45.

Boon, M.A., Furr, H.L. and Hausman, A. (2001) Using Internet technology to request travel informationand purchase travel services: a comparison of x’ers, boomers and mature market segments. In:Manzanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) Consumer Psychology of Tourism,Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK, pp. 187–193.

Cooper, C.P. (1981) Spatial and temporal patterns of tourist behaviour. Regional Studies 15, 359–371.Crotts, J.C. and Reid, L.J. (1993) Segmenting the visitor market by the timing of their activity decisions.

Visions In Leisure and Business 12, 4–7.Decrop, A. (1999) Tourist’s decision-making and behavior processes. In: Pizam, A. and Mansfeld, Y. (eds)

Consumer Behavior in Travel and Tourism. The Haworth Hospitality Press, Binghampton, New York,pp. 103–133.

Elliott, R. and Hamilton, E. (1991) Consumer choice tactics and leisure activities. International Journal ofAdvertising 10, 325–332.

Engel, J.F., Kollat, D.T. and Blackwell, R.D. (1968) Consumer Behaviour. Holt, Reinhart & Wilson, New York.Fesenmaier, D. and Jeng, J. (2000) Assessing structure in the pleasure trip planning process. Tourism

Analysis 5, 13–27.Holbrook, M.B. and Hirschman, E.C. (1982) The experiential aspects of consumption: consumer fantasies,

feelings, and fun. Journal of Consumer Research 9, 132–140.Hyde, K.F. (2000) A hedonic perspective on independent vacation planning. In: Woodside, A.G., Crouch,

G.I., Manzanec, J.A., Oppermann, M. and Sakai, M.Y. (eds) Consumer Psychology of Tourism, Hospitalityand Leisure. CAB International, Wallingford, UK, pp. 177–191.

Kemperman, A., Borgers, A., Oppewal, H. and Timmermans, H. (1997) Visitors’ activity patterns in atheme park. Paper presented at the Battle For The Tourist Conference, Eindhoven.

King, R.L. and Woodside, A.G. (2001) Qualitative comparative analysis of travel and tourism purchase-con-sumption systems. In: Manzanec, J.A., Crouch, G.I., Ritchie, J.R.B. and Woodside, A.G. (eds) ConsumerPsychology of Tourism, Hospitality And Leisure, Vol. 2. CAB International, Wallingford, UK, pp. 67–86.


Lue, C., Crompton, J.L. and Fesenmaier, D.R. (1993) Conceptualization of multi-destination pleasure trips.Annals of Tourism Research 20, 289–301.

Mergen, B. (1986) Travel as play. In: Mergen, B. (ed.) Cultural Dimensions of Play, Games, and Sport. GeorgeWashington University, Washington, DC, pp. 103–111.

Moutinho, L. (1987) Consumer behaviour in tourism. European Journal of Marketing 21, 2–44.Oppermann, M. (1994) Length of stay and spatial distribution. Annals of Tourism Research 21, 834–836.Oppermann, M. (1997) First-time and repeat visitors to New Zealand. Tourism Management 18, 177–181.Ryan, C. (2002) ‘The time of our lives’ or time for our lives: an examination of time in holidaying. In:

Ryan, C. (ed.) The Tourist Experience, 2nd edn. Continuum, London, pp. 201–212.Um, S. and Crompton, J.L. (1990) Attitude determinants in tourism destination choice. Annals of Tourism

Research 17, 443–448.van Raaij, W.F. and Francken, D.A. (1984) Vacation decisions, activities, and satisfactions. Annals of Tourism

Research 11, 101–112.Woodside, A.G. and Lysonski, S. (1989) A general model of traveler destination choice. Journal of Travel

Research 27, 8–14.Woodside, A.G. and Macdonald, R. (1994) General systems framework for customer choice processes of

tourism services. In: Gasser, G. and Weiermair, K. (eds) Spoilt For Choice. Kultur Verlag Thaur,Germany.

Woodside, A.G. and Dubelaar, C. (2002) A general theory of tourism consumption systems: a conceptualframework and an empirical exploration. Journal of Travel Research 41, 120–132.

168 K.F. Hyde


Chapter fourteenA Model of Vacation Choice: an Integration of Personalityand Vacation Choice with Leisure Constraints Theory

Robyn L. McGuigganSydney Graduate School of Management, Parramatta, NSW 2150, Australia

Abstract

Over the years, a number of tourist choice models have been advanced by researchers but, in the study oftourist choice, two areas in particular appear to be lacking – the possible role of the individual differencevariable, personality, in that choice, and the importance of constraints during the choice process. Thischapter presents a model of vacation choice based on the constraints model of leisure participation pro-posed by Crawford et al. (Leisure Sciences 13, 309–320, 1991) and modified by Jackson et al. (Leisure Sciences15, 1–11, 1993), but incorporating the individual difference variable, personality. In building on Jackson etal.’s model, it is proposed that personality will influence vacation preference through the development ofmotives, and that personality will influence the perception of both intrapersonal constraints and assess-ment of the ability to negotiate them. It is further proposed that an individual’s weighted motives will giverise to weighted vacation attribute preferences and that successful negotiation of interpersonal constraintswill lead to jointly agreed weighted vacation attribute preferences.

Introduction

There can be little doubt as to the importanceof the tourism industry to Australia. Tourismis recognized as the country’s largest industryand the major producer of export income, aswell as the greatest generator of employment(Fiszman, 2000). In 1999–2000, tourism repre-sented 14.1% of Australia’s total export earn-ings and comprised 61.9% of services exports(Bureau of Tourism Research, 2000). During1997–1998, tourism expenditure directly con-tributed $25.2 billion or 4.5% to Australia’sgross domestic product. ‘Holidays’ (as in

travel for leisure), is the most popular reasongiven for travel to and from Australia. For theyear ended June 2000, 55.7% of inbound trav-ellers and 48% of domestic travellers quoted‘holidays’ as their main purpose of journey(Bureau of Tourism Research, 2000). Holidaysor vacations are a highly regarded leisureactivity and have become a cultural norm, notonly for Australian society, but also for manysocieties around the world (Kelly andFreysinger, 2000).

Considering the size and growth rate ofthe tourism industry and the importanceplaced on touristic activities, the tourism mar-



ket presents substantial opportunities for fur-ther development. By improving their under-standing of tourist behaviour, marketers willbecome more effective in serving the needs,and meeting the expectations, of tourists(Morrison et al., 1994).

For a marketer in tourism, understandingwhy people decide to travel and influences ontheir choice of destination are of critical impor-tance in developing appropriate marketingstrategies. Woodside and Carr (1988, p. 2) notethat ‘most consumers making vacation plans“actively consider” (evaluate the pros and consof) about four vacation destination choices’.How are these destinations selected for evalua-tion? To complicate matters further, consumersmay perceive the same destination as offeringtotally different experiences. For example, onetourist may see a trip to Vanuatu as offering atranquil environment in which to relax, donothing and recuperate from work, whereasanother may envisage active water sports suchas sailing, snorkelling and scuba diving. In mar-keting a destination, it is important to under-stand customers’ vacation needs andpreferences and their underlying influences.

Although a number of tourist choice mod-els have been advanced by researchers over theyears (e.g. Goodrich, 1978; Moutinho, 1987;Woodside and Lysonski, 1989; Haider andEwing, 1990; Sirakaya et al., 1996; Moutinho,2000), two specific areas of research have beenneglected in the study of tourist choice:

● the role that the individual difference vari-able – personality – might play in thatchoice; and

● the role of constraints during the choiceprocess.

The objective of this chapter is to presenta model of vacation choice based on the con-straints model of leisure participation devel-oped by Crawford et al. (1991) and modifiedby Jackson et al. (1993), but which also incor-porates the individual difference variable,personality.

Personality and Vacation Choice

With the consumption of tourism productsbeing highly individualized, a clear under-

standing of consumer behaviour and psychol-ogy is fundamental if businesses are todevelop products, services and promotionsthat satisfy the needs of their chosen market(Dickman, 1999). While there are a numberof psychological factors that may be consid-ered as objects of study, personality is one ofthe better known and, potentially, the mostuseful concept (Jackson et al., 1999).Although personality research has gainedprominence within the field of leisure overthe past two decades, relatively little tourismresearch has focused on examining the per-sonality of tourists.

Plog (1972) was the first to conductresearch on personality as applied specificallyto tourists (Madrigal, 1995). He sought tolink personality traits directly with touristbehaviour and distributed tourists over cate-gories along a bell curve, from psychocentricsto allocentrics. According to his model, psy-chocentrics are persons who are less adven-turous, inward looking, who prefer familiarityin their surroundings and concentrate onpopular destinations such as Coney Islandand Miami Beach (Plog, 1972). Allocentrics,at the other end of the continuum, are adven-turous and prepared to take risks and fre-quent destinations which appeal to theirsense of adventure and novelty, such as Africa(Sharpley, 1994). Plog’s model is the mostwell-known and applied personality constructwithin the field of tourism. Although somesupport for Plog’s model, particularly in rela-tion to locus of control, has been forthcom-ing (Nickerson and Ellis, 1991; Griffith andAlbanese, 1996), other researchers havefound little to support his model (e.g. Lee-Hoxter and Lester, 1988; Smith, 1990;Madrigal, 1995; Jackson et al., 1999). Thus theability of Plog’s model to adequately explaintourist behaviour has not been proved and itsusefulness in adding to the current under-standing of tourists is questionable.

In recognition of the shortcomings ofPlog’s model, a small number of researchershave adopted a more conservative approachin their investigations of the tourist–person-ality relationship, utilizing better-known per-sonality measures. For example, Gilchrist etal. (1995) investigated the relationshipbetween sensation-seeking and adventure

170 R.L. McGuiggan


travel. They observed that adventure trav-ellers scored significantly higher onZuckerman’s (1991) sensation-seeking scale.Frew and Shaw (1997, 1998) were able todemonstrate a link between personality andtourist attractions using Holland’s (1985)theory of personality types. They found thatof the 93 combinations of personality andtourist attractions examined, between one-third and one-half showed a significant rela-tionship. Holbrook and Olney (1995) wereable to demonstrate a relationship betweenthe personality variable of romanticism andwanderlust, with the preference for morerisky vacations and warmer climates.

Although studies on personality andtourist destination choice are limited,authors such as Fridgen (1996) argue thatpersonality manifests itself in the travel deci-sions tourists make and their choices whileon vacation. He suggests, for example, thatan introverted person would not enjoy avacation requiring extensive contact withstrangers, while someone with a more conser-vative personality is unlikely to choose anadventure holiday (pp. 58–59).

The relationship between personality andleisure choice has been extensively studied,and there is strong evidence that a relation-ship exists (Furnham and Heaven, 1999).Recent studies in leisure research, which havedemonstrated strong correlations betweenpersonality variables and leisure, have doneso by addressing a number of the drawbacksin previous studies (McGuiggan, 1996, 2000).In particular, they have addressed the need tomeasure variables at the same level of abstrac-tion, measuring activity attributes rather thanthe activity itself. They have also taken intoconsideration a number of other variablesthat may intervene between personality andleisure choice and so have measured leisurepreference rather than actual choice. Sincerecreational tourism is a leisure activity (Iso-Ahola, 1983), it should follow that if the sameissues are addressed in the study of tourismchoice, such a relationship should also existbetween personality and tourism preference.

Therefore in the study of the relationshipbetween personality and tourism choice, itcan be argued that the following should beconsidered in the model:

● tourist destinations should be measured interms of attributes, rather than actual des-tinations;

● personality should be a predictor of touristpreference, rather than choice; and

● the role of other constraining variablesneeds to be considered in the relationship.

Constraints Theory

Although little research has been undertakeninvolving the role of constraints in touristchoice, a considerable body of literature hasappeared since 1985 investigating constraintson leisure participation. Initially researchersequated constraints with barriers to participa-tion (Crawford and Godbey, 1987; Bialeschkiand Henderson, 1988); more recent researchhas argued that constraints do not necessarilyprohibit total engagement in an activity.Rather, constraints should be seen as modifiersof choice and participation, resulting in eithernon-participation or some revision of partici-pation – for example, participation to a lesserextent or participation in a different activity(Crawford et al., 1991; Jackson, 1993; Jacksonand Rucks, 1995). It should also be noted thatconstraints might be perceived, rather thanfactual (Henderson et al., 1988), or not be per-ceived, but still exist (Jackson, 1997).

A number of elements of leisure con-straints have emerged from the literature thatmay be relevant in tourist choice. These are:

● Constraints can be classified into threetypes – intrapersonal, interpersonal andstructural (Crawford et al., 1991) – thesemay affect preference as well as participa-tion (Crawford and Godbey, 1987). Whileall studies appear to support the existenceof these three categories of constraints,individual authors disagree on the classifi-cation of individual constraints. For exam-ple, Boothby et al. (1981) argue that timeconstraints are personal constraints,whereas most other researchers treat timeas an external, structural constraint(Crawford et al., 1991).

● There is a hierarchy of constraints fromintrapersonal to interpersonal to struc-tural (Crawford et al., 1991), but Hudson(1998) argues that, at least in the case ofsnow skiing, this hierarchy may not hold.

A Model of Vacation Choice 171


● Constraints can vary in intensity and notall constraints will result in non-participa-tion; rather participation will depend onthe motivation and ability of the individualto negotiate through them (Jackson andRucks, 1995).

Crawford et al.’s (1991) hierarchical modelof leisure constraints is depicted in Fig. 14.1.

Intrapersonal constraints

Intrapersonal constraints have been conceptu-alized as being the first level of the hierarchyand influence preference, rather than inter-vening between preference and choice(Crawford et al., 1991). They are defined byCrawford and Godbey (1987, p. 122) as ‘indi-vidual psychological states and attributes’ – forexample, stress, depression, religiosity, anxiety,perceived self-skill and subjective evaluations ofappropriateness of various activities. It is con-tended that if these intrapersonal constraintsare too great, then no motivation to participatewill eventuate. This does not imply that an indi-vidual does not have a need/want/interest toundertake an activity, but rather that there isinsufficient motivation to progress further inthe decision making process as the constraintson participation are already too great. Absenceof these constraints, or successful negotiationof them, will lead to the development of prefer-ences (Raymore et al., 1993).

Interpersonal constraints

Interpersonal constraints ‘involve the interac-tions and relationships between individuals’(Samdahl and Jekubovich, 1997, p. 431). Theinability to find a partner or friend with

whom to engage in an activity would act as aninterpersonal constraint, as would the non-compatibility of preferences, or the co-partici-pant’s structural constraints. Interpersonalconstraints are seen as interacting with pref-erence for, and also participation in, activitiesundertaken with others (Raymore et al.,1993). If these constraints cannot be success-fully negotiated then participation will notoccur. If successful negotiation can occur,then a final level of constraints will be facedbefore participation is ensured.

Structural constraints

Structural constraints are those that occurbetween preference and participation(Crawford and Godbey, 1987, p. 123). Theseconstraints could involve factors such asfinancial resources, availability of time, familylife cycle stage, season or climate, referencegroup attitudes as to appropriateness of theactivity, lack of opportunity or access. It isreported that the absence of, or the success-ful negotiation of, these constraints willensure participation.

Hierarchy of constraints

A number of authors have supported thebreakdown of constraints into the three cate-gories described above (e.g. Jackson et al.,1993; Raymore et al., 1993; Samdahl andJekubovich, 1997; Hudson, 1998). AlthoughRaymore et al.’s (1993) study provided strongsupport for the order of the hierarchy pro-posed by Crawford et al. (1991), other studieshave not been able to replicate the hierarchy(Samdahl and Jekubovich, 1997; Hudson,

172 R.L. McGuiggan

Intrapersonalconstraints

Interpersonalconstraints

Structuralconstraints

Leisurepreferences

Participation (ornon-participation)

Interpersonalcompatibility

and coordination

Fig. 14.1. A model of leisure constraints (Crawford et al., 1991).


1998). Samdahl and Jekubovich’s (1997)qualitative study provides some support forthe hierarchical perspective, but they reportthat the evidence for the hierarchy is notabsolute. For example, one of their respon-dents utilized an interpersonal relationship toovercome an intrapersonal constraint – lackof self-confidence in dining alone. Theirargument is flawed – it could be argued that‘eating out’ is regarded as a group activity inwestern culture and the respondent did notexhibit a lack of self-confidence in diningwith others, only in dining alone. Thereforethere was no intrapersonal constraint inhibit-ing her from dining with others, only inter-personal – identification of a diningcompanion. Furthermore, Jackson et al.(1993) do not expound that an intrapersonalconstraint will invariably lead to non-partici-pation, but can result in a negotiationprocess, which may lead to modification ofthe behaviour. So if eating out alone is theinitial goal, then negotiation at the intraper-sonal constraint level could lead to modifica-tion of the behaviour – eat out with a friend.

The results of Hudson’s (1998) study areconsistent with the first level in the hierarchy;he found intrapersonal constraints can deter-mine preference. He found no evidence thatpeople confront interpersonal constraintsbefore structural constraints when decidingon a skiing holiday. He concluded that inter-personal constraints were the least evident inhis study. Perhaps this is because his studyfocused on only one activity – skiing. Hefound that skiers appeared to have negotiatedboth intra- and interpersonal constraints, andtheir focus was mainly on structural con-straints. Conversely, non-skiers scored highon intrapersonal constraints and showed apreference for not skiing. Their lack of inter-est in interpersonal constraints related to ski-ing were possibly due to lack of motivation inundertaking the activity.

Negotiation of constraints

Jackson et al. (1993) state explicitly that peo-ple do not react passively to constraints ontheir leisure, but actively negotiate to con-tinue to participate, albeit in a modified form

from that originally intended. He states that‘participation is dependent not on theabsence of constraints (although this may betrue for some people) but on negotiationthrough them’ (p. 4) leading to modificationrather than non-participation in activities.This is supported by findings in some studiesthat constraints did not lead to a decrease inleisure participation (Kay and Jackson, 1991;Shaw et al., 1991). In their paper investigatingthe leisure activities of high school students,Jackson and Rucks (1995) suggest that thereare three broad categories of negotiationadopted by people – cognitive adaptation (e.g.change leisure preference) or behaviouralmodification, either in terms of the leisureactivity (e.g. playing basketball in the backyardinstead of elsewhere) or in their non-leisuredomain (e.g. modifying non-leisure activitiesto fit timing of leisure activity).

Jackson et al. (1993) introduce the idea oflevel of motivation into the model, agreeingwith Wright and Goodale (1991) that peoplewho do not report a desire to begin, orincrease participation in an activity are in facta heterogeneous group. They either have nointerest in the activity or are currently partici-pating to their ideal level; alternatively theyhave been unable to successfully negotiate anantecedent constraint. They suggest in theirmodified model (Fig. 14.2) that the relativestrength of a person’s motivation to partici-pate, together with the various constraints,will determine the extent to which they willbe willing to negotiate.

Jackson et al. (1993) also suggest that innegotiating an antecedent constraint, peoplewill look forward to other constraints theyexpect to encounter, both interpersonal andstructural, and assess their ability to negotiatethese. This will affect the manner in whichthey will negotiate the original antecedentconstraint. They also suggest that if at a laterstage in the decision process an unanticipatedinterpersonal or structural constraint appearsand cannot be successfully negotiated, then afeedback loop to an earlier stage in the modelwill occur. For example, if a structural con-straint is encountered, such as lack of availabil-ity of an airfare for a planned vacation, thedecision maker may return to discussions withthe accompanying parties, or they may reassess



their own preferences in the light of this newconstraint. Therefore, although Jackson et al.(1993) support the concept of a leisure con-straint hierarchy, they do not regard it as anabsolute, thus allowing the decision maker torevert to an earlier level of constraints if a laterconstraint cannot be successfully negotiated.

Proposed Model of Tourist Choice

Figure 14.3 presents the author’s conceptual-ization of how leisure constraints theorycould be integrated into a model of personal-ity, vacation preference and choice. Theabove discussion is drawn upon in validatingthe model. Arguments are also presented tojustify the modification.

The model is based on Jackson et al.’s(1993) model (Fig. 14.2), and includes thefeedback loops to earlier stages of the modelif constraints are not successfully negotiated,or alternatively, there is lack of interest incontinuing the decision making process, asdiscussed in their paper but not depicted intheir model. As these relationships have beenpreviously discussed in this chapter, there willbe no further discussion here. The inclusionof personality in the model has also been dis-cussed and there will be no further elabora-tion. The remainder of the chapter will focuson the justification for the additional rela-tionships proposed and the modification ofexisting components of the model. The fol-lowing modifications will be discussed andresearch propositions identified:

● the influence of personality on intraper-sonal constraints;

● the role of personality in determining therelative weighting of motives;

● the relationship between weighted motivesand weighted individual vacation attributepreferences;

● the role of interpersonal constraints indetermining the weighted interpersonalvacation attribute preferences.

Influence of personality on intrapersonalconstraints

As indicated above, intrapersonal constraintsinvolve ‘individual psychological states andattributes’ (Crawford et al., 1991, p. 311) andthese may be actual or perceived. In theirpaper, Crawford et al. specify a number ofintrapersonal constraints that, it could beargued, are influenced by personality. Forexample, people low in dogmatism, high ininnovativeness and high in optimum stimula-tion level might be less likely to see lack ofprior experience as a constraint. Other-directed consumers might be more con-cerned about reference group attitudes to theappropriateness of specific activities or desti-nations. Personality variables, such as need forcognition, may affect the type of negotiationutilized, or the assessment of ability to success-fully negotiate constraints. In support of thisargument, Courneya and Hellsten (1998)found that neuroticism positively correlated,and conscientiousness negatively correlated,

174 R.L. McGuiggan

Intrapersonalconstraints

Interpersonalconstraints

Structuralconstraints

Leisurepreference

Participation (ornon-participation)

Interpersonalcompatibility

and coordination

Motivations(attractions)

Fig. 14.2. Leisure participation: a balance between constraints and motivation (Jackson et al., 1993).



Per

sona

lity

Wei

ghte

dm

otiv

es

Wei

ghte

d in

divi

dual

vaca

tion

attr

ibut

epr

efer

ence

s

Wei

ghte

d in

terp

erso

nal

vaca

tion

attr

ibut

epr

efer

ence

s

Intr

aper

sona

lco

nstr

aint

sIn

terp

erso

nal

cons

trai

nts

Lack

of i

nter

est

Fee

dbac

k lo

op

Str

uctu

ral

cons

trai

nts

No

vaca

tion

Vac

atio

nch

oice

Unsuccessfulnegotiation



Fig.

14.

3.M

odel

of v

acat

ion

choi

ce: a

n in

tegr

atio

n of

per

sona

lity

and

v aca

tion

choi

ce w

ith le

isur

e co

nstr

aint

s th

eory

.


with exercise barriers such as lack of energyand embarrassment. Furthermore, they foundan inverse relationship between extraversionand the exercise barrier of lack of energy.

Proposition 1: Personality will influence theperception of both intrapersonal con-straints and assessment of ability to negoti-ate them.

Influence of personality on motives

Many researchers see motives as a major deter-minant of tourist behaviour. The distinctionbetween personality and motives in determin-ing preference or choice is not always clarifiedin the tourism and leisure literature, nor in thewider consumer behaviour literature. Otherauthors, such as Maddi (1980), draw a clear dis-tinction between the two dimensions. Evidencedoes exist that these concepts are different,and that personality influences motives. Forexample, Costa and McCrae (1988) undertooka study to determine the correlation betweenthe Neo Personality Inventory (the Big Five)and Murray’s list of 20 manifest needs. Theirstudy clearly showed the two constructs to bedifferent; e.g. finding that neuroticism was cor-related to the need for defensiveness, help,protection and social acceptance. They con-cluded that trait psychologists ‘should considerthe explicitly motivational aspects of their con-structs’ (p. 264) and that motivational invento-ries should not be seen as a substitute forbroader based personality instruments.Mooradian and Oliver (1996) also used the BigFive model of personality and showed a clearrelationship between the personality dimen-sions and shopping motives.

In terms of leisure research, Iwasaki andMannell (1999) have shown a strong relation-ship between the Intrinsic Leisure Motivationpersonality scale and intrinsic motivation toplay a puzzle game. Furthermore, Courneyaand Hellsten (1998) found that each dimen-sion of the Big Five correlated with exercisemotives. Neuroticism correlated with physicalappearance and weight control, and extraver-sion with the need for socialization and meet-ing people. Extraversion and conscientiousnesscorrelated with the need to maintain fitnessand health, extraversion and openness to expe-

rience with the need to maintain mental healthand stress relief, and extraversion, openness toexperience and neuroticism with the need forfun and enjoyment.

The previous discussion of Plog’s modelsuggests a relationship exists between person-ality and tourist needs. Psychocentrics have aneed for stability and familiarity, while allo-centrics demonstrate need for adventure andnovelty. Although the literature is somewhatconfusing in terms of the distinction betweenpersonality and motives, arguments can bemade that the two concepts are discrete andthat personality influences needs. Perhapsconsideration of motives as a moderating vari-able between personality and tourist prefer-ence will increase the predictive capacity ofthe model.

Proposition 2: Personality influences vaca-tion preference through the formation ofmotives.

Motives and individual vacation attributepreferences

Since the 1970s, tourism researchers haveconcentrated on establishing ‘push factors’(motives) – e.g., need to escape and need forsocializing – and the investigation of theirrelationship to ‘pull factors’ (attributes of thevarious destinations, such as infrastructure,scenic attractions and scenery, accessibilityand historical interest) (Sirakaya et al., 1996).If a tourist destination wishes to attract allo-centrics, with their need for activity and nov-elty, the destination should offer variety fromeveryday living, a large range of activitiesfrom which to choose and a non-touristyatmosphere (Ross, 1998). Thus the need forstimulation (or in fact the opposite) willinfluence preference in vacation attributes.Taking another example, a tourist with aneed to be seen and recognized (social sta-tus) may have a preference for a destinationthat is expensive and prestigious (Fridgen,1996). Decrop (1999) has suggested thatthere are two levels of push factors (motives)for tourism emerging from the literature andhis data. Firstly there is the base need tobreak from routine – escaping both tempo-rally (every day routine) and spatially

176 R.L. McGuiggan


(home). He suggests the second level of‘push factors’ is more specific and will varyfrom individual to individual. In his study heidentifies six of these factors, but many otherauthors have identified various sets ofmotives (Gibson and Yiannakis, 2002).

How can it be explained that, despite thepresence of constraints, some people will stillparticipate in an unchanged activity, otherswill modify their participation in some way,while a third group will decide not to partici-pate (Kay and Jackson, 1991)? Strength ofmotivation could provide a partial answer tothis question. If the person’s motivation totake a vacation is not particularly strong, thepersonal cost of negotiating any constraintmay appear too difficult and so participationwill not occur. At the other extreme, if themotivation is particularly strong, a personmay be willing to participate in an activityregardless of the constraints encountered.Altered participation may be due to moderatelevels of motivation leading an individual tonegotiate to some extent on an acceptabletype of vacation. As suggested by a number ofauthors (e.g. Woodside and Jacobs, 1985; Pittsand Woodside, 1986; Fodness, 1994), peoplemay have a number of vacation needs thatvary in relative importance. These in turnwould be expected to manifest in terms ofpreferences for various vacation attributeswith a corresponding level of importance. Forexample, relaxation may be an extremelystrong motive for a particular individual, lead-ing to a strong preference for a ‘kids club’.On the other hand, the preference for adven-ture may not be as strong, so although thismotive translates into a preference for abseil-ing and parasailing, this is more likely to benegotiable than the stronger preference for a‘kids club’.

Proposition 3: An individual’s weightedmotives will give rise to weighted vacationattribute preferences.

Interpersonal constraints and interpersonalvacation attribute preferences

Vacations in general tend to be a social activ-ity – people go on holiday with family orfriends, or join tour groups. Therefore it

might be expected that interpersonal con-straints – the identification of appropriatetravel companions or the incompatibility ofpersonal vacation attribute weightings –would be of major import in the ultimatechoice of vacation destination. As indicatedin Crawford et al.’s (1991) model (Fig. 14.1),the ability to successfully negotiate interper-sonal constraints to achieve interpersonalcompatibility and coordination is essential toparticipation in an activity. The only differ-ence in the current model is that achieve-ment of this ‘compatibility and coordination’will result in jointly agreed weighted vacationattribute preferences. The initial strength andoverlap of the individuals’ attribute prefer-ences will determine any compromises andthe relative importance of particular agreedattribute preferences. In their study, Samdahland Jekubovich (1997) found that peopleoften compromise activities to fit in withleisure partners – indicating that their socialmotives were not negotiable, but activitieswere. The inability to successfully negotiatethese interpersonal constraints will result ineither lack of interest in taking the vacationor a feedback loop to reassess individual vaca-tion attribute preferences.

Proposition 4: Successful negotiation ofinterpersonal constraints will lead tojointly agreed weighted vacation attributepreferences.

It should be borne in mind that actualchoice will depend on the match between theattributes of the vacation destination and theinterpersonal vacation attribute preferenceweightings – provided structural constraintsdo not exist, or have been successfully negoti-ated. As identified by Stemerding et al.(1996), destination attributes may be ‘rejec-tion-inducing’, where consumers utilize anon-compensatory decision rule – a destina-tion does not have an attribute that is rated asessential in the interpersonal attribute prefer-ences. Other attributes Stemerding et al.(1996) suggest may be thought of as ‘trade-off attributes’, where importance or attrac-tiveness weightings need to be assigned to theindividual vacation destination attributes(Sirakaya et al., 1996, pp. 62–63) – consumersuse a compensatory decision rule.



Discussion

The model presented in this chapter couldexplain why studies investigating the correla-tion between personality variables and touristchoice have been less than impressive.Personality constructs are valid in the studyof tourist choice, but may affect behaviourdifferently, depending on the various con-straints encountered. For example, healthissues, an intrapersonal constraint, may havea significant modifying effect on the prefer-ences of someone with an adventurous, risk-taking personality. Similarly a person with anintroverted personality who wants to holidaywith a spouse who has an extraverted person-ality – an interpersonal constraint – mayrequire negotiation of joint vacation prefer-ences that are quite different from the intro-vert’s individual preferences. Structuralconstraints, for example, non-availability offlights within a timeframe imposed by familycommitments, may mean negotiation andselection of a non-optimal destination, thusleading to a lower correlation between per-sonality and actual choice.

There are a number of challenges to befaced in researching the proposed model.Firstly, the choice of personality theory ortraits that would most likely be associated withvacation needs must be addressed.Considering the dearth of studies in this area,perhaps borrowing from the leisure literaturemight prove fruitful. Secondly, intrapersonal,interpersonal and structural constraints inrelation to vacation choice need to be estab-lished. It is suggested that qualitative researchmethods would be most effective in elicitingthese and again the leisure literature may pre-sent a starting point for this research. Themeasurement of vacation attributes is morestraightforward – experts could be used torate destinations in terms of the variousattributes described in the tourism literature.Alternatively, the methodology used byStemerding et al. (1996), involving a modifiedRepertory Grid technique, could be applied.Rather than using randomly generated triads,

experimental design principles were used togenerate the groupings for attribute elicita-tion. Respondents were asked to identify theleast and the most preferred destination andto identify the attributes leading to this evalu-ation.

Establishing individual motives, vacationattribute preferences and their relativestrengths, as well as the interpersonal vaca-tion preferences and their relative strengths,presents the greatest challenge. Qualitativeresearch methods will need to be utilized, atleast initially, in a longitudinal study.Individuals who are planning to go on vaca-tion need to be identified and interviewedbefore having chosen a destination or havingbeen on vacation confounds their memory ofthe decision process. It remains to be seenwhether individuals can differentiate betweentheir own individual vacation attribute prefer-ence strengths and their negotiated interper-sonal vacation attribute preference strengths.At this point intrapersonal and interpersonalconstraints would need to be identified. A fol-low-up interview would be required to deter-mine their actual choice and any structuralconstraints encountered. The feedback loopswould need to be probed during both inter-views as well as, perhaps, the negotiationstrategies used.

Conclusion

Although as yet untested, this model offersplausible explanations as to why higher cor-relations have not been found between per-sonality and tourist choice. The argumentcould be extended to many other choice situ-ations. Investigation of the model offers thepossibility of providing a more in-depthunderstanding of the role of personality,motives and constraints in the choice oftourist vacations. This should in turn providetourist operators with a better insight intohow their customers choose between alterna-tives, and allow the development of superior,targeted marketing strategies.

178 R.L. McGuiggan


References

Bialeschki, M.D. and Henderson, K.A. (1988) Constraints to trail use. Journal of Park and RecreationAdministration 6, 20–28.

Boothby, J., Tungatt, M.F. and Townsend, A.R. (1981) Ceasing participation in sports activity: reportedreasons and their implications. Journal of Leisure Research 13, 1–14.

Bureau of Tourism Research (2000) Impact – a monthly facts sheet on the economic impact of tourism and the latestvisitor arrivals trends. www.tourism.gov.au/publications/impact/html (last accessed 15 October, 2002).

Cost, P.T.J. and McCrae, R.R. (1988) From catalog to classification: Murray’s needs and the five-factormodel. Journal of Personality and Social Psychology 55, 258–265.

Courney, K.S. and Hellsten, L.-A.M. (1998) Personality correlates of exercise behavior, motives, barriers andpreferences: an application of the five-factor model. Personality and Individual Differences 24, 625–633.

Crawford, D.W. and Godbey, G. (1987) Reconceptualising barriers to family leisure. Leisure Sciences 9, 119–127.Crawford, D.W., Jackson, E.L. and Godbey, G. (1991) A hierarchical model of leisure constraints. Leisure

Sciences 13, 309–320.Decrop, A. (1999) Personal aspects of vacationers’ decision making processes: an interpretivist approach.

Journal of Travel and Tourism Marketing 8, 59–68.Dickman, S. (1999) Tourism and Hospitality Marketing. Oxford University Press, South Melbourne, Victoria.Fiszman, S. (2000) Cashing in on golden legacy. The Daily Telegraph, Sydney, p. 18.Fodness, D. (1994) Measuring tourist motivation. Annals of Tourism Research 21, 555–581.Frew, E.A. and Shaw, R.N. (1997) Personality, career choice and tourism behaviour: an exploratory study.

In: Tourism Research Building a Better Industry, Australian Tourism and Hospitality Research Conference.Bureau of Tourism Research, Canberra, pp. 275–287.

Frew, E.A. and Shaw, R.N. (1998) The relationship between personality, gender, and tourism behaviour. In:Eighth Australian Tourism and Hospitality Research Conference. Bureau of Tourism Research, Canberra,pp. 82–94.

Fridgen, J. (1996) Dimensions of Tourism. Educational Institute of the American Hotel and MotelAssociation, East Lansing, Michigan.

Furnham, A. and Heaven, P. (1999) Personality and Social Behaviour. Oxford University Press, London.Gibson, H. and Yiannakis, A. (2002) Tourist roles: needs and the lifecourse. Annals of Tourism Research 29,

358–383.Gilchrist, H., Povey, R., Dickinson, A. and Povey, R. (1995) The sensation-seeking scale: its use in a study of the

characteristics of people choosing ‘adventure holidays’. Personality and Individual Differences 19, 513–516.Goodrich, J.N. (1978) The relationship between preferences for and perceptions of vacation destinations:

application of a choice model. Journal of Travel Research 16, 8–13.Griffith, D.A. and Albanese, P.J. (1996) An examination of Plog’s psychographic travel model within a stu-

dent population. Journal of Travel Research 34, 47–51.Haider, W. and Ewing, G.O. (1990) A model of tourist choices of hypothetical Caribbean destinations.

Leisure Sciences 12, 33–47.Henderson, K.A., Stalnaker, D. and Taylor, G. (1988) The relationship between barriers to recreation and

gender role personality traits for women. Journal of Leisure Research 20, 69–80.Holbrook, M.B. and Olney, T.J. (1995) Romanticism and wanderlust: an effect of personality on consumer

preferences. Psychology & Marketing 12, 207–222.Holland, J.L. (1985) Making Vocational Choices: a Theory of Vocational Personalities and Work Environments, 2nd

edn. Prentice Hall, Englewood Cliffs, New Jersey.Hudson, S. (1998) An extension of constraints theory related to the consumer behaviour of skiing. PhD,

University of Surrey, UK.Iso-Ahola, S.E. (1983) Towards a social psychological theory of recreational travel. Leisure Studies 2, 45–56.Iwasaki, Y. and Mannell, R.C. (1999) Situational and personality influences on intrinsically motivated

leisure behavior: interaction effects and cognitive processes. Leisure Sciences 21, 287–306.Jackson, E.L. (1993) Recognizing patterns of leisure constraints: results from alternate analyses. Journal of

Leisure Research 25, 129–149.Jackson, E.L. (1997) In the eye of the beholder: a comment on Samdahl & Jekubovich (1997), ‘A critique of

leisure constraints: Comparative analysis and understandings.’ Journal of Leisure Research 29, 458–469.



Jackson, E.L. and Rucks, V.C. (1995) Negotiation of leisure constraints by junior-high and high-school stu-dents: an exploratory study. Journal of Leisure Research 27, 85–105.

Jackson, E.L., Crawford, D.W. and Godbey, G. (1993) Negotiation of leisure constraints. Leisure Sciences 15,1–11.

Jackson, M.S., Schmierer, C.L. and White, G.N. (1999) Is there a unique tourist personality which is predic-tive of tourist behaviour? In: Ninth Australian Tourism and Hospitality Research Conference. Bureau ofTourism Research, Canberra, pp. 29–47.

Kay, T. and Jackson, G. (1991) Leisure despite constraint: the impact of leisure contraints on leisure partic-ipation. Journal of Leisure Research 23, 301–313.

Kelly, J.R. and Freysinger, V.J. (2000) 21st Century Leisure. Allyn and Bacon, Boston, Massachusetts.Lee-Hoxter, A. and Lester, D. (1988) Tourist behavior and personality. Personality and Individual Differences

9, 177–178.Maddi, S.R. (1980) Personality Theories: a Comparative Analysis, 4th edn. Dorsey Press, Homewood, Illinois.Madrigal, R. (1995) Personal values, traveler personality type, and leisure travel style. Journal of Leisure

Research 27, 125–142.McGuiggan, R.L. (1996) The relationship between personality, as measured by the Myers–Briggs Type

Indicator, and leisure preferences. PhD dissertation, Macquarie University, Sydney, Australia.McGuiggan, R.L. (2000) The Myers–Briggs Type Indicator and leisure attribute preference. In: Woodside,

A.G., Crouch, G.I., Mazanec, J.A., Oppermann, M. and Sakai, M.Y. (eds) Consumer Psychology ofTourism, Hospitality and Leisure. CAB International, Wallingford, UK, pp. 245–267.

Mooradian, T. and Oliver, J. (1996) Shopping motives and the five-factor model. Psychological Reports 78,579–592.

Morrison, A.M., Hsieh, S. and O’Leary, J. (1994) Segmenting the Australian domestic travel market by hol-iday activity participation. The Journal of Tourism Studies 5, 39–56.

Moutinho, L. (1987) Consumer behavior in tourism. European Journal of Marketing 21, 1–44.Moutinho, L. (2000) Consumer behaviour. In: Moutinho, L. (ed.) Strategic Management in Tourism. CAB

International, Wallingford, UK, pp. 41–78.Nickerson, N.P. and Ellis, G.D. (1991) Traveler types and activation theory: a comparison of two models.

Journal of Travel Research 29, 26–31.Pitts, R.E. and Woodside, A.G. (1986) Personal values and travel decisions. Journal of Travel Research 25,

20–25.Plog, S.C. (1972) Why destinations rise and fall in popularity. In: Richie, J.R.B. and Goeldner, C.R. (eds)

Details restated in Understanding Psychographics in Tourism Research 1987 Travel, Tourism and HospitalityResearch. John Wiley & Sons, New York, pp. 203–213.

Raymore, L., Godbey, G., Crawfarod, D.W. and von Eye, A. (1993) Nature and process of leisure con-straints: an empirical test. Leisure Sciences 15, 99–113.

Ross, G.F. (1998) The Psychology of Tourism, 2nd edn. Hospitality Press Pty Ltd, Melbourne.Samdahl, D.M. and Jekubovich, N.J. (1997) A critique of leisure constraints: comparative analysis and

understandings. Journal of Leisure Research 29, 430–453.Sharpley, R. (1994) Tourism, Tourists and Society. ELM Publications, Huntingdon, UK.Shaw, S.M., Bonen, A. and McCabe, J.F. (1991) Do more constraints mean less leisure? Examining the rela-

tionship between constraints and participation. Journal of Leisure Research 23, 286–300.Sirakaya, E., McLellan, R.W. and Uysal, M. (1996) Modeling vacation destination decisions: a behavioral

approach. Journal of Travel and Tourism Marketing 5, 57–75.Smith, S.L.J. (1990) A test of Plog’s allocentric/psychocentric model: evidence from seven nations. Journal

of Travel Research 28, 40–43.Stemerding, M.P., Oppewal, H., Beckers, T.A.M. and Timmermans, H.J.P. (1996) Leisure market segmenta-

tion: an integrated preferences/constraints-based approach. Journal of Travel and Tourism Marketing 5,161–185.

Woodside, A.G. and Carr, J.A. (1988) Consumer decision making and competitive marketing strategies:applications for tourism planning. Journal of Travel Research 26, 2–7.

Woodside, A.G. and Jacobs, L.W. (1985) Step two in benefit segmentation: learning the benefits realized bymajor travel markets. Journal of Travel Research 24, 7–13.

Woodside, A.G. and Lysonski, S. (1989) A general model of traveler destination choice. Journal of TravelResearch 27, 8–14.

Wright, B.A. and Goodale, T.L. (1991) Beyond non-participation: validation of interest and frequency ofparticipation categories in constraints research. Journal of Leisure Research 23, 314–331.

Zuckerman, M. (1991) Psychobiology of Personality. Cambridge University Press, Cambridge.

180 R.L. McGuiggan


Chapter fifteenEffects of Holiday Packaging on Tourist Decision Making:Some Preliminary Results

Walaiporn Rewtrakunphaiboon1 and Harmen Oppewal21School of Management, University of Surrey, Guildford, Surrey GU2 7XH, UK; 2Departmentof Marketing, Monash University, PO Box 197, Caulfield East, Victoria 3145, Australia

Abstract

This study aims to investigate how packaging can be used for destination marketing. The focus is on howbundling and presentation format of package information influence consideration and intention to visit.Most tourism researchers have ignored the fact that destination choice is actually a complex and multifac-eted decision process in which destinations are selected in combination with other elements that comprisea holiday. Tourism literature also seems to assume that consideration sets are likely to be stable over timeand that destination marketers should try to get their destinations into the consumers’ consideration setsas early as possible. The present study will investigate these assumptions by looking into the effects of holi-day packaging on consideration and intention to visit. Preliminary results suggest that package informa-tion format has an impact on intention to visit.

Introduction

Package holidays have expanded enormouslyin the tourism industry over the past decadeand the competition in this industry has beenvery intensive (Taylor, 1998). Package holi-days comprise over half of all overseas holidayvisits by UK residents and the share of pack-age holidays has increased from 51.9% in1996 to 53.4% in 2001 (National Statistics,2002). Despite this increase, the effects ofholiday packaging on consideration andintention to visit have received very littleattention in tourism research. A better under-

standing of these effects will be beneficial fordestination marketers, travel agents as well astour operators.

Although tourism researchers have pro-posed several models of the destinationchoice process, the effects of marketing stim-uli such as package holidays have hardly beeninvestigated. Studies on tourist decision mak-ing have traditionally focused on destinationchoice alone and have ignored the possibleeffects of packaging. Other authors, however,have explained that many travel decisions arenot single independent choices of separateelements such as destination, accommodation



or transportation but rather are a complexset of multi-faceted decisions in which thechoices for different elements are interre-lated (Dellaert et al., 1998). Burkart (1984)argues that package holidays are productsthat are destination indifferent and empha-sizes that destination is no longer a prime fac-tor in the choice of a holiday. Studies ondestination choice alone are therefore lim-ited and may be misleading if destinationchoice is related to the choice of other com-ponents of the package (Dellaert et al., 1997).

Holiday packaging is a form of bundling(e.g. Suri and Monroe, 1995; Soman andGourville, 2001). Research on bundling hasmostly taken an economic perspective, whichfocuses on the change in profits, and con-sumer surplus that ensues if bundles areoffered. Only few studies have explored con-sumer behaviour in response to bundling(Harlam et al., 1995; Suri and Monroe, 1995).Furthermore, the impacts of package holidayson consideration and intention to visit havehardly been examined in tourism research.

The present study focuses on the effects ofbundling and the presentation of packageinformation on consideration and intentionto visit. This chapter will first review relevantliterature and propose hypotheses regardingpotential effects of bundling and informationformat on consideration and intention tovisit. It will then explain how we propose totest these ideas and present some preliminaryresults from the first data collection. Thechapter will end with a conclusion section.

Literature Review

Destination choice process

In tourism literature, the notion of choicesets has been widely accepted as a usefulstructural framework for conceptualizing howtourists sift through a large number of vaca-tion destinations available to them(Crompton, 1992; Crompton and Ankomah,1993). The concept postulates that there is afunnelling process, which involves a relativelylarge initial set of destinations being reducedto a smaller late set, from which a final desti-nation is selected (Ankomah et al., 1996).

Choice sets are most likely to be applicablewhen the purchase task is a new or modifiedone in which individuals typically seek infor-mation and evaluate alternatives, and whenthe purchase entails some degree of high risk(Spiggle and Sewall, 1987). Many vacationdestination choice decisions are likely to meetthese two criteria (Crompton, 1992).

There seems to be a general agreementamong tourism researchers that destinationselection goes through three major stages: (i)development of an early consideration setwhich has been generally called awarenessset; (ii) a discarding of most of destinations toform a smaller late consideration set orevoked set; and (iii) a final destination choice(Crompton and Ankomah, 1993). Awarenessset refers to all the destinations of which anindividual may be aware at any given time.Early or initial consideration set refers to thedestinations which a traveller is consideringas possible vacation destinations within someperiod. Late consideration set refers to thedestinations which a traveller is consideringas probable destinations within some periodof time (Crompton, 1992, pp. 423–424).Evoked set refers to the destinations that theconsumer is aware of and has some likelihoodgreater than zero of visiting within some timeperiod (Woodside and Sherrell, 1977, p. 15).Crompton (1992) explains that, conceptually,a key differentiating element between earlyand late consideration sets is a period of timeelapsing between them that is sufficientlylong to enable individuals to evaluate andreduce their list of destinations from a broadset of possible destinations to a narrower setof probable destinations.

The early consideration set consists ofthose destinations considered by a potentialtourist to be possible for a vacation. If a desti-nation is not in an individual’s early consider-ation set, then it has no chance of beingselected (Crompton and Ankomah, 1993).However, the previous claim has hardly beenempirically tested. Furthermore, it has beendisputed that consideration set formation isactually a dynamic process that may evolveuntil consumers decide to make a final choice(Shocker et al., 1991; Mitra, 1995).Alternatives can be dropped from and addedto the consideration set as it develops

182 W. Rewtrakunphaiboon and H. Oppewal


(Klenosky and Rethans, 1988). In an unfamil-iar choice situation, consideration sets aremore likely to be unstable over time. Even inhighly familiar choice situations, the contentsof the consideration set will vary across differ-ent purchase situations (Klenosky andRethans, 1988). These unstable conditions ofthe consideration sets across occasions woulddepend upon the factors that are present atthe time of decision making (Mitra, 1995) andindividual factors (Shocker et al., 1991). It isargued here that it may also depend uponwhether alternatives are bundled and howinformation about these bundles is presented.

Bundling

Bundling has become a common phenome-non in marketing. Package holidays consti-tute a form of bundling. Bundling is definedas the marketing of two or more productsand/or services in a single ‘package’ for aspecial price (Guiltinan, 1987, p. 74). Thereare two main forms of bundling. Purebundling refers to the products or servicesavailable only in package form. Mixedbundling refers to the products or servicesavailable individually or as a package (Adamsand Yellen, 1976).

Package holidays are an important compo-nent of the travel business. A package holidayis a combination of many components of avacation such as transportation, accommoda-tion, sightseeing and meals, which are sold toconsumers at a single price (Bojamic andCalantone, 1990). Package holidays varyaccording to their inclusiveness. Among allpackage holidays, the simplest is the basicpackage holiday, which typically includestransportation and accommodation only.Inclusive package holidays also offer somesightseeing and entertainment at the destina-tion. All inclusive package holidays includemeals and are sometimes escorted (Sheldonand Mak, 1987).

Consumers may not know the prices of theindividual components because they pur-chase the entire package either from a travelagent who is a retailer of vacation products orfrom a tour operator who creates the pack-age, and publishes and then distributes the

brochure. Convenience and cheaper priceare the two reasons why tourists purchasepackage holidays (Sheldon and Mak, 1987).

The price of a bundle is usually lower thanit would be if the items were purchased differ-ently (Schwartz and Cohen, 1999). Con-sumers may evaluate the worth of a productdifferently when it is in a bundle than when itis not. Linking two or more items together islikely to influence the context in which con-sumers evaluate those items because it will lit-erally force the consumer to evaluate them inthe context of one another (Harris, 1997).Furthermore, the consumer may save timeand effort, including the cost of gatheringinformation, by buying complementary com-modities in a single package (Paroush andPeles, 1981; Oppewal and Holyoake, 2003).

One item in the bundle may be already inan individual’s consideration set, but otheritems in the bundle often are not (Suri andMonroe, 1995). Bundling can educate as wellas remind consumers. Value for money andquality of the product or services help con-sumers to distinguish an alternative and makeit more or less probable that the alternativewill enter the awareness set and the consider-ation sets of those consumers who find thefeatures attractive for their own purposes(Shocker et al., 1991). Price bundling oftenincreases purchase likelihood (Soman andGourville, 2001). Furthermore, past researchhas shown that a bundle was evaluated morefavourably and chosen more often when itscomponents were presented in segregated(separate price tags) versus consolidated (sin-gle, equivalent tag) fashion (Yadav andMonroe, 1993; Suri and Monroe, 1995).

Leisure travellers are likely to have a posi-tive evaluation towards destinations in theirevoked sets (Woodside and Sherrell, 1977;Woodside and Ronkainen, 1980). Althoughthey associate each destination with a particu-lar set of benefits, they may be willing to sub-stitute one benefit–destination combinationwith a competing benefit–destination pack-age (Woodside and Carr, 1988). One particu-lar condition in which this may occur is whenthe destination is part of a bundle. Othercomponents in the bundle may assist a lesspreferred destination to enter the considera-tion set although it was not initially included.

Effects of Holiday Packaging on Tourist Decision Making 183


For example, a traveller may think of going toMalaysia when asked directly. When there is acompetitive discount package that has otherdestinations such as Thailand, the same per-son may choose Thailand and hence end upvisiting Thailand after all.

Information format

An issue that is related to bundling is the for-mat in which product information is presented.Information format refers to the presentationand organization of information about theavailable alternatives and their attributes(Cooper-Martin, 1993, p. 240). Informationformat affects the way consumers process thatinformation (Bettman and Kakkar, 1977).Previous research suggests that the way a prod-uct is presented influences the importanceconsumers assign to various attributes whenmaking a purchase decision. There are two pos-sible ways in which information format canaffect consumer perception. Firstly, the percep-tual salience of an attribute increases whenproducts are displayed according to a specificattribute, for example, by brand or by price(Tversky, 1969; Glass and Holyoak, 1986). Thisincreases the importance that attributes receivewhen the consumers evaluate products and/ormake purchase decisions (MacKenzie, 1986;Hutchinson and Alba, 1991). Secondly, pre-senting product information according to agiven attribute makes it easier for consumers tocompare alternatives using that particularattribute (Russo, 1977; Bettman, 1979).

Results of a study on wine purchasing byAreni et al. (1999) reveal that categorizingwine by region of origin increases the salienceof region and makes it easier for consumers tocompare the alternatives on the region of ori-gin attribute. Alternatives originating from anunfavourable wine region, such as Texas, werepenalized by this decision criterion and salesdropped when Texan wines were on promo-tion. Organizing products according to dis-tinct levels or values of a specific attributehence affects purchase likelihood. It affectsthe perceptual salience of that attribute andthe ease with which it can be used to makecomparisons. In a related paper, Areni (1999)therefore suggests that products from lesser-

known regions should highlight attributeswith which they are likely to compare favour-ably (e.g. value for money, variety) rather thanfocusing solely on the region.

Similarly, products of equivalent value maybe evaluated differently depending on theway in which the bundle is presented to theconsumers (Harlam et al., 1995). Della Bitta etal. (1981) studied consumer perception ofcomparative price advertisements and foundthat presenting information as different com-binations of sale price, regular price, percent-age off and dollar amount off resulted indifferent perceptions of the offer. Simonsonand Winer (1992) also find that there is a sig-nificant interaction between display formatand task condition. They suggest that chang-ing product display format can influence con-sumer purchases.

Past research has consistently demon-strated that consumer evaluations of and pref-erences for a product can be influenced bythe context of the choice task, including theset of alternatives considered and the mannerin which these alternatives are presented (e.g.Simonson and Tversky, 1992). Conceptualinsights and empirical findings will be usefulto understand how a bundle should be pre-sented and how specific consumer character-istics may affect consumer evaluation of abundle (Krish et al., 1994).

Hypotheses

The results of the previous studies can beapplied to tourism products. Package holi-days are mostly presented with destination asthe ‘heading’ that labels the package. Ifanother attribute, price, is used to label apackage holiday, the likelihood that a particu-lar destination is selected may change. Itseems that this possible effect has not yetbeen tested in tourism research.

Based on the literature review, we proposethe following two hypotheses:

H1: Bundling influences (i) probability ofentering the consideration set and (ii) intentionto visit. H2: The format in which attribute informationis presented affects (i) probability of entering theconsideration set and (ii) intention to visit.



Method

To test the hypotheses, an experimental ques-tionnaire concerning overseas beach holidayswas developed. Two hundred students tookpart in this experiment. The questionnaireintroduced participants to a hypothetical situ-ation in which they had to assume that theyhad won an overseas beach holiday voucherworth £1000. They must spend this voucheron flights and accommodation for two per-sons. They were free to take anyone with themand the travel companion would be happy togo to any destination. If they did not spend allof £1000, they would be given the remainderas a voucher for an overseas holiday next year.

Design

Participants were randomly assigned to oneof three different questionnaire conditions:(i) unbundled condition, (ii) bundle condi-tion I (destination as a heading) or (iii) bun-dle condition II (price as a heading). Theunbundled condition was a control conditionin which respondents were only asked to ratetheir intention to visit for a list of destinationsand select the destinations that they wouldconsider visiting from this same list. Afterthat, they were asked to separately indicatetheir preferences for other attributes. Thebundle conditions were the experimentalconditions, in which respondents received atreatment, a series of package holidays, whichwere designed by using the design principlesapplied in conjoint analysis. Attributesdescribed in the package holidays includeddestination, package price, type of accommo-dation, number of nights and name of travelagent. In bundle condition I, the packageholiday presented the name of destination asa heading. In bundle condition II, the pack-age holiday presented price as the heading.

Dependent variables

There were two main dependent variables inthe study: probability of entering the consid-eration set and intention to visit. These twovariables were measured before and after the

treatment. For intention to visit, respondentswere asked to rate for each destination howlikely they were to visit this destination on aseven-point scale anchored by will definitelynot visit (1) to will definitely visit (7). Forprobability of entering the consideration set,they were asked to select the destinationsthat they would consider visiting from a listof destinations.

Procedure

First, subjects in all conditions were askedabout their preferred type of holiday, pre-ferred month and their overseas holidayexperience including beach holidays. Next,they were exposed to the hypothetical situa-tion. Then, they were asked to rate intentionto visit and select the destinations that theywould consider visiting for their next overseasbeach holidays from a list.

Next, only the respondents assigned to thebundle conditions were exposed to one of thetwo experimental package holiday conditions:package holidays with destination as a head-ing or package holidays with price as a head-ing. To help induce the experimentaltreatments and to allow manipulation checks,respondents in the experimental conditionswere asked to rate the attractiveness of thepackage holidays by indicating how attractivethey found each package holiday on a seven-point scale anchored by not at all attractive(1) and very attractive (7).

After the treatment, only subjects in thebundle conditions were asked to further ratedestinations on their intention to visit and toselect destinations that they would consider vis-iting for their next overseas beach holidaysfrom a list that included partly similar, partlynew destinations. This procedure imple-mented an additional control condition, i.e.we observed ratings for the destinations thatthe respondents had previously seen and thosethat they had not seen in the earlier tasks.

Analysis and Results

Preliminary results in this section concern theintention to visit ratings that were obtained



for two destinations, Turkey and Tunisia,across three different conditions. The analysiswas performed on the differences in intentionto visit between the control condition and thetwo experimental conditions. In the controlcondition, respondents did not see any bun-dle (unbundled condition) while the respon-dents in the two experimental conditions wereexposed to either package holidays with thedestination as a heading (bundle condition I)or package holidays with price as a heading(bundle condition II). For the two experimen-tal conditions, the intention to visit ratingscollected after exposure to the treatment wereused for analysis. The presence of the two des-tinations in the package holidays was systemat-ically varied. However, our focus here is onthe analysis of intention to visit across thethree conditions.

The first analysis was performed on inten-tion to visit Turkey. Figure 15.1 illustratesmean of intention to visit Turkey across thethree conditions. It shows the actual meanscores of the three conditions as follows: bun-dle condition I (mean = 3.77, SD = 1.75), bun-dle condition II (mean = 4.33, SD = 1.70) andunbundled condition (mean = 3.95, SD =1.72). To test the hypotheses, an ANOVA withplanned contrasts was conducted. The effectof package holidays on intention to visitappeared as not significant for Turkey(F(1,193) = 0.108, n.s.).

The contrast representing the differencebetween bundle condition I (destination as a

heading) and bundle condition II (price as aheading) was, however, significant for Turkey(F(1,193) = 4.285, P = 0.040). These resultssuggest that the format in which package hol-idays were presented influenced intention tovisit Turkey. When respondents were asked torate their intention to visit Turkey withoutany exposure to any package holidays(unbundled condition), the mean scores ofintention to visit were 3.95. The mean scoresdropped to 3.77 when Turkey was presentedas a heading (bundle condition I). In con-trast, the mean scores increased to 4.33 whenprice was presented as a heading (bundlecondition II). This suggests that Turkey wasperceived as a less preferred destination inwhich presenting package holidays with desti-nation as a heading decreased the intentionto visit, whereas presenting package holidayswith price as a heading increased the inten-tion to visit.

Similar results also occurred to the otherholiday destination in our study, Tunisia.Figure 15.2 shows mean intention to visitTunisia across the three conditions. ForTunisia, the mean scores of intention to visit inthe unbundled condition were 4.28 (SD =1.62). The mean scores dropped to 3.85 (SD =1.62) when Tunisia was presented as the head-ing (bundle condition I) whereas the meanscores increased to 4.42 (SD = 1.62) when pricewas presented as the heading (bundle condi-tion II). The contrast representing the differ-ence between the bundle conditions and the


Condition

UnbundledBundle IIBundle I

Mea

n of

inte

ntio

n to

vis

it

4.4

4.3

4.2

4.1

4.0

3.9

3.8

3.7

3.6

Fig. 15.1. Mean of intention to visit Turkey, bybundle condition.

Condition

UnbundledBundle IIBundle I

Mea

n of

inte

ntio

n to

vis

it

4.5

4.4

4.3

4.2

4.1

4.0

3.9

3.8

3.7

Fig. 15.2. Mean of intention to visit Tunisia, bybundle condition.


unbundled condition was not significant at theconventional alpha level of 5% (F(1,195) =0.265, n.s.). Therefore, H1 is not supported bythe results of two holiday destinations.

However, the contrast between the twobundle conditions was statistically significant(F(1,195) = 4.981, P = 0.027). The meanscores of intention to visit to Tunisia in theunbundled condition were 4.28. The meanscores dropped to 3.85 when Tunisia was pre-sented as a heading (bundle condition I),whereas the mean scores increased to 4.42when price was presented as a heading (bun-dle condition II). Therefore, H2 is supportedby the results of both holiday destinations.

Conclusion

Existing tourism literature on destinationchoice suggests that it is almost impossible fordestinations that are not included in the earlyconsideration set to enter the considerationset and be a potential candidate for choice.However, this chapter argued that holidaypackaging could assist destinations to enterthe consideration set and hence increase theintention to visit. Research carried out to testthis assumption will be useful to understandtourist decision making as well as to supportfurther use and development of packagingstrategies to market holiday destinations. The

present chapter presented hypotheses,methodology and some preliminary results ofan experimental study that aims to investigatethese possible effects.

Further analysis is required before any sub-stantial conclusions can be drawn. However,our preliminary results for two destinationssuggest that package information format doeshave an impact on intention to visit. Theresults of Turkey and Tunisia showed that ifprice is used as the heading of package holi-days, the intention to visit increases relativelyto a control condition in which no packageholiday information is provided. In contrast,intention to visit decreases when destination ispresented as the heading of a package holiday.

Another possible explanation for thisresult is that presenting package holidays withdestination as the heading increases thesalience of destination as an attribute. Asshown in a study by Areni et al. (1999) in thecontext of wine marketing, presenting destina-tion as a heading would be a disadvantage forless popular destinations such as Tunisia andTurkey. Presenting destination as a headinghighlights the attribute on which these coun-tries compete unfavourably with other destina-tions. Our results so far suggest that lesspreferred destinations should be presented ina format with price as a heading rather thandestination as a heading. Further analyses andtests on this assumption will soon be reported.


References

Adams, W.J. and Yellen, J.L. (1976) Commodity bundling and the burden of monopoly. Quarterly Journal ofEconomics 90, 475–498.

Ankomah, P.K., Crompton, J.L. and Baker, D. (1996) Influences of cognitive distance in vacation choice.Annals of Tourism Research 23, 138–150.

Areni, C.S. (1999) An examination of the impact of product organization and region equity on the com-parison and selection of wines. Advances in Consumer Research 26, 359–364.

Areni, C.S., Duhan, D.F. and Kiecker, P. (1999) Point-of-purchase displays, product organization and pur-chase likelihoods. Journal of the Academy of Marketing Science 27, 428–441.

Bettman, J.R. (1979) An Information Processing Theory of Consumer Choice. Addison-Wesley, Reading,Massachusetts.

Bettman, J.R. and Kakkar, P. (1977) Effects of information presentation format on consumer informationacquisition strategies. Journal of Consumer Research 3, 233–240.

Bojamic, D.C. and Calantone, R.J. (1990) A contribution approach to price bundling in tourism. Annals ofTourism Research 17, 528–540.

Burkart, A.J. (1984) Marketing package holidays. Service Industries Journal 4, 187–192.Cooper-Martin, E. (1993) Effects of information format and similarity among alternatives on consumer

choice processes. Journal of Academy of Marketing Science 21, 239–246.Crompton, J. (1992) Structure of vacation destination. Annals of Tourism Research 19, 420–434.Crompton, J.L. and Ankomah, P.K. (1993) Choice set propositions in destination decisions. Annals of

Tourism Research 20, 461–476.


Della Bitta, A.J., Monroe, K.B. and McGinnis, J.M. (1981) Consumer perceptions of comparative priceadvertisements. Journal of Marketing Research 18, 416–427.

Dellaert, B.G.C., Borgers, A.W.J. and Timmermans, H.J.P. (1997) Conjoint models of tourist portfoliochoice: theory and illustration. Leisure Sciences 19, 31–58.

Dellaert, B.G.C., Ettema, D.F. and Lindh, C. (1998) Multi-faceted tourist travel decisions: a constraint-based conceptual framework to describe tourists’ sequential choices of travel components. TourismManagement 19, 313–320.

Glass, A.L. and Holyoak, K.J. (1986) Cognition. Random House, New York.Guiltinan, J.P. (1987) The price bundling of services: a normative framework. Journal of Marketing 51,

74–85.Harlam, B.A., Krishna, A., Lehmann, D.R. and Mela, C. (1995) Impact of bundle type, price framing and

familiarity on purchase intention for the bundle. Journal of Business Research 33, 57–66.Harris, J. (1997) The effects of promotional bundling on consumers’ evaluations of product quality and

risk of purchase. Advances in Consumer Research 24, 168–172.Hutchinson, J.W. and Alba, J.W. (1991) Ignoring irrelevant information: situational determinants of con-

sumer learning. Journal of Consumer Research 18, 346–357. Klenosky, D.B. and Rethans, A.J. (1988) The formation of consumer choice sets: a longitudinal investiga-

tion at the product class level. Advances in Consumer Research 15, 13–18.Krish, R., Paul, P. and Srivastava, J. (1994) Behavioral perspectives on bundling research. Advances in

Consumer Research 21, 255.MacKenzie, S.B. (1986) The role of attention in mediating the effect of advertising on attribute impor-

tance. Journal of Consumer Research 13, 174–195. Mitra, A. (1995) Advertising and the stability of consideration sets over multiple purchase occasions.

International Journal of Research in Marketing 12, 81–94. National Statistics (2002) Travel Trends 2001. The Stationery Office, London.Oppewal, H. and Holyoake, B. (2003) Bundling and retail agglomeration effects on shopping behavior.

Journal of Retail and Consumer Services (in press).Paroush, J. and Peles, Y. (1981) A combined monopoly and optimal packaging. European Economic Review

15, 373–383.Russo, J.W. (1977) The value of unit price information. Journal of Marketing Research 14, 193–201.Schwartz, Z. and Cohen, E. (1999) The perceived value of value meals: an experimental investigation into

product bundling and decoy pricing in restaurant menus. Journal of Restaurant and FoodserviceMarketing 3, 19–33.

Sheldon, P.J. and Mak, J. (1987) The demand for package tours: a mode choice model. Journal of TravelResearch 25, 13–17.

Shocker, A.D., Ben-Akiva, M., Boccara, B. and Nedungadi, P. (1991) Consideration set influences on con-sumer decision-making and choice: issues, models, and suggestions. Marketing Letters 2, 181–197.

Simonson, I. and Tversky, A. (1992) Choice in context: tradeoff contrast and extremeness aversion. Journalof Marketing Research 29, 281–295.

Simonson, I. and Winer, R.S. (1992) The influence of purchase quantity and display format on consumerpreference for variety. Journal of Consumer Research 19, 133–138.

Soman, D. and Gourville, J.T. (2001) Transaction decoupling: how price bundling affects the decision toconsume. Journal of Marketing Research 38, 30–44.

Spiggle, S. and Sewall, M.A. (1987) A choice sets model of retail selection. Journal of Marketing 51, 97–111.Suri, R. and Monroe, K.B. (1995) Effect of consumers’ purchase plans on the evaluation of bundle offers.

Advances in Consumer Research 22, 588–593.Taylor, P. (1998) Mixed strategy pricing behaviour in the UK package tour industry. International Journal of

the Economics of Business 5, 29–46.Tversky, A. (1969) Intransitivity of preferences. Psychological Review 76, 31–48. Woodside, A.G. and Carr, J.A. (1988) Consumer decision making and competitive marketing strategies:

applications for tourism planning. Journal of Travel Research 26, 2–7.Woodside, A.G. and Ronkainen, I.A. (1980) Tourism management strategies for competitive vacation desti-

nations. In: Hawkins, D.E., Shafter, E.L. and Rovelstad, J.M. (eds) Tourism Marketing and ManagementIssues. George Washington University, Washington, DC, pp. 3–19.

Woodside, A.G. and Sherrell, D. (1977) Traveler evoked, inept and inert sets of vacation destinations.Journal of Travel Research 16, 14–18.

Yadav, M.S. and Monroe, K.B. (1993) How buyers perceive savings in a bundle price: an examination of abundle’s transaction value. Journal of Marketing Research 30, 350–358.



Chapter sixteenAn Examination of the Antecedents and Consequences ofCustomer Satisfaction

Yuksel Ekinci1 and Ercan Sirakaya2

1School of Management, University of Surrey, Guildford GU2 7XH, UK; 2Texas A&MUniversity, 256A Francis Hall, 2261 TAMU, College Station, TX 77843-2261, USA

Abstract

This study investigates the relationships between customer satisfaction, service quality and overall attitude.To this end, two conceptual frameworks and ten hypotheses are tested using structural equation model-ling. The data are collected in a restaurant setting using a convenience sampling procedure. The findingsindicate that the evaluation of service quality leads to customer satisfaction, and satisfaction rather thanservice quality is a better reflection of overall attitudes. Also, desires congruence and ideal self-congruenceare found to be antecedents of customer satisfaction.

Introduction

The research on customer satisfaction has along history that dates back to the early1960s. Since then more than 15,000 tradeand academic papers have been published(Peterson and Wilson, 1992). However,despite the growing interest in customer sat-isfaction, it still remains an elusive conceptdue to a number of theoretical and method-ological shortcomings that continue to per-sist in the literature. At the heart of them arethe antecedents and consequences of cus-tomer satisfaction. In particular, examina-tions of the relationship between customersatisfaction and theoretically related variablessuch as attitude and service quality have pro-duced controversial results and therefore it

has been subject to hefty debates (Ekinci andRiley, 1998; Fournier and Mick, 1999).

Though helpful, these debates havecaused confusion in both the service qualityand satisfaction literature. For example,Oliver (1980, 1997) argued that customer sat-isfaction is a similar construct to attitudes.According to his postulation, customer satis-faction mediates changes between pre-pur-chase and post-purchase attitudes. Hence,customer satisfaction is dynamic and quicklydecays into one’s attitudes. However, in thequality literature, the concept of service qual-ity is substituted by customer satisfactionwhile proposing exactly the same type of rela-tionship. Parasuraman et al. (1988) arguedthat service quality is more universal andenduring and therefore can be a better



reflection of an attitude. Furthermore, theauthors claimed that customer satisfaction isspecific to a service encounter and anantecedent of service quality (Parasuraman etal., 1994). The literature is awash withdetailed arguments of this kind but the out-come of this research is inconclusive.

Despite the above studies that offerinsight into the relationships between cus-tomer satisfaction, service quality and atti-tudes, a holistic conceptual framework is stillmissing. Theoretical arguments suggest thateither customer satisfaction or service qual-ity is similar to an attitude, but fail to pro-vide empirical evidence. Hence, the role ofattitude in the formation of satisfaction andevaluation of service quality remainsunclear. On the other hand, there areempirical studies that investigate the rela-tionship between service quality and cus-tomer satisfaction but they are limited inquantity. Most of them have produced mixedresults and therefore the relationshipbetween the two concepts is left to theresearchers’ own interpretation.

Basically, three types of conclusions aredrawn from these studies (Ekinci and Riley,1998). The first one suggests that an evalua-tion of customer satisfaction leads to servicequality whereas the second one suggests thatan evaluation of service quality leads to cus-tomer satisfaction. It is difficult to determinethe exact nature of relationship from thesestudies, the last one rejects both formulationand argues that the two concepts, servicequality and customer satisfaction are thesame, and that there is no need to make a dis-tinction between the two through a causalrelationship. While the literature on cus-tomer satisfaction and service quality progressin parallel, the fact that research into theactual differences between the two conceptswould be mutually beneficial and should berecognized.

The purpose of this study is to examinethe relationship between customer satisfac-tion and the other theoretically related vari-ables: service quality, attitudes, self-conceptcongruence, desires congruence and behav-ioural intentions. To this end, we developedten hypotheses and then tested two compet-ing models.

Background: Antecedents andConsequences of Customer Satisfaction

Customer satisfaction

The definition of satisfaction has shown greatdiversity within industry and societal perspec-tives. Among the ten proposed theories ‘theexpectancy disconfirmation theory’ has beenthe most popular one due to its broadlyapplicable conceptualization (Oh and Parks,1997). This theory suggests that satisfaction isrelated to the size and direction of the dis-confirmation experience that occurs as resultof comparing service performance againstexpectations (Oliver, 1980).

Despite the popularity of the disconfirma-tion theory, it suffers from its simplicity. Someof the empirical studies using this paradigmfailed to explain satisfaction judgement in dif-ferent consumption situations. Mittal et al.(1998) argued that the relationship betweenattribute-level performance and overall satis-faction changes marginally (diminishing sen-sitivity for both negative and positiveperformance) rather than linearly and sym-metrically. Other scholars emphasized thatthe satisfaction process is more complex thanis explained by the disconfirmation theory(LaTour and Peat, 1979; Oliver, 1980;Churchill and Suprenant, 1982). Oliver(1997, p. 13) offered an updated definitionthat reflects the findings of recent theoreticaland empirical studies.

Satisfaction is the consumer’s fulfilmentresponse. It is a judgement that a product orservice feature, or the product or service itself,provided (or is providing) a pleasurable level ofconsumption-related fulfilment, including levelsof under- or overfulfilment.

According to this definition, the fulfilmentresponse is a pleasurable state that is derivedby reducing the pain when a problem issolved or alleviated. However, pleasure can beobtained not only by the unexpected effect ofoverfulfilment, but also underfulfilment suchas when the actual damage is less thanexpected. Oliver (1997) argued that satisfac-tion is strongly related with fulfilling needsbut this notion requires more elaboration indifferent consumption situations. The abovedefinitions promote two notions. Firstly, satis-

190 Y. Ekinci and E. Sirakaya


faction is the result of direct experience withproducts or services and secondly, it occurs bycomparing this experience against a standard(e.g. expectations). Oliver (1980) furtherexplained how a satisfaction judgement isaccumulated during the consumption period.Figure 16.1 shows the cognitive process of sat-isfaction formation and its relationship withother constructs.

According to Fig. 16.1, a customerapproaches the service encounter with anantecedent attitude (ATTa) which might havebeen accumulated through previous experi-ences, word of mouth communications ormarketing promotions before purchasing(time 1 = t1). The antecedent attitude is afunction of expectation. The intention to pur-chase behaviour at the pre-consumptionperiod is influenced by the ATTa. During theconsumption period, the customer compareshis expectations with the service perfor-mance. By the same token, a disconfirmationprocess occurs at this stage. The outcome ofthis can be positive, negative or neutral.Hence, a satisfaction decision begins toemerge during the consumption period andbecomes dominant towards the end of thisperiod. In line with this, a satisfaction decisionis a function of expectations and the level ofthe disconfirmation experience. However, thissatisfaction decision is time and situation spe-cific, and, therefore, soon decays into ATTa toestablish continuous attitudes (ATTc). Here,satisfaction acts as a moderating variable

between ATTa and ATTc. Therefore, thedirection and magnitude of satisfaction servesas an input to form the ATTc, which has beenadopted at the post-consumption period. Thelatter attitude influences the customer’s inten-tion to re-purchase at the post consumption(time 2 = t2). The ATTc is then a function ofATTa and satisfaction whereas the intention tore-purchase (t2) is a function of the previousintention to purchase (t1), satisfaction and theATTc. The following sets of expressions sum-marize these relationships.

ATTa(t1) = f (expectations)intention (t1) = f (ATTc(t1))satisfaction = f (expectations, disconfir-

mation)ATTc(t2) = f (ATTa (t1), satisfaction)intention (t2) = f (intention (t1), satisfac-

tion, ATTc(t2))

Oliver’s (1997) conceptualization isnotable as it illustrates both the cognitiveprocesses of satisfaction formation and itsrelationship with other constructs, in particu-lar, the intention to purchase and attitudestowards a product. The discussion leads tothe following two hypotheses:

H1: Customer satisfaction has a positive associ-ation with behavioural intention (recommendand return).H2: Customer satisfaction has a positiveassociation with attitudes towards a serviceorganization.

Antecedents and Consequences of Customer Satisfaction 191

b1e1:

Expectation

ATTITUDE(a)

Intention

Satisfaction

ATTITUDE(c)

Intention

Disconfirmation

Disconfirmationperiod

t1 t2

Pre-consumption Consumption Post-consumption

Fig. 16.1. The process of satisfaction formation. Adapted from Oliver (1980), p. 465.


Attitudes

According to the most frequently cited defini-tion by Allport (1935), attitudes are learnedpredispositions to respond to an object orclass of objects in a consistently favourable orunfavourable way. The ‘theory of reasonedaction’ is the most prominent model thatexplains consumer attitudes towards anaction through behavioural intentions (Ajzenand Fishbein, 1980). According to this model,an attitude consists of three elements: (i) thenet outcome of performing the behaviour(e.g. beliefs on the costs and benefits of thisbehaviour such as visiting a country); (ii)social pressure or subjective norm (the influ-ence of other people); and (iii) the perceivedbehavioural control (the extent to which aperson believes he/she has control over per-formance of the behaviour). These threefunctions could be assessed simultaneously bydirectly asking the importance of a bundle ofattributes representing beliefs.

Attitudes towards purchase behaviour areunderlined by many factors. Although a num-ber of functional theories of attitudes havebeen developed, the one proposed by Katz(1960) has perhaps received the most atten-tion. According to his theory, there are fourfunctions of attitudes known as underlyingmotivations: the utilitarian, the ego-defensive,the knowledge and the value-expressive functions.

The utilitarian function of attitudes refersto the fact that people tend to acquire atti-tudes because they desire certain outcomes.For example, a positive attitude towards acampus restaurant may be developed becauseit offers a convenient location. The ego-defensive function of attitudes may be heldbecause it allows people to protect themselvesfrom being exposed of their weaknesses.Hence, people tend to hide their inadequaciesfrom the harsh realities of the external world.For example, consumers may hold positive atti-tudes towards diet products or dandruff-freeshampoos to defend themselves against anunderlying feeling of physical inadequacy.

The value-expressive function of attitudesallows people to express their central valuesor self-concept. In many ways, this is the com-plete opposite of the ego-defensive function.

For example, a conservative person may holda positive attitude towards British Airways as itrepresents being British. Maoi and Olson(1994) showed that people with value-expres-sive attitudes have significant relationsbetween value importance and their attitudesor behaviour, whereas people with utilitarianattitudes do not.

The knowledge function of attitudes mayserve as a standard since it helps us to under-stand our universe. By the same token, such anevaluation is cognitive and it attaches meaningto the self and its relation to environment.Maoi and Olson (1994, p. 301) stated ‘to someextents, the knowledge function may exists inall attitudes as they serve to organise informa-tion about attitude objects’. In general, there isample evidence showing that attitudes influ-ence consumer behaviour (Burnkrant andPage, 1982). As consumers bring their atti-tudes with them to the service encounter, theyalso use them for the evaluation of services.Hence customer satisfaction influences contin-uous attitude (ATTc) at the post-consumptionphase; however, before that happens theantecedent attitude (ATTa) also influencescustomer satisfaction. Therefore, the relation-ship between the two concepts is bi-direc-tional. We argue that this is an important pathand was not specified by Oliver (1980) in hismodel. Thus, the following hypotheses havebeen proposed to guide this study.

H3: Customers’ favourable attitude towards aservice organization has a positive associationwith customer satisfaction.H4: Customers’ favourable attitude towards aservice organization has a positive associationwith behavioural intention.

Service quality

Definitions of quality have varied over theyears. Early definitions suggest that qualityshould be seen as conformance to specifica-tions. Hence, positive quality is obtainedwhen the product matches with predeter-mined standards or specifications. However,this is considered as being manufacture-ori-ented and therefore many scholars arguedthat service quality should be customer-ori-ented (Reeves and Bednar, 1994).



Consequently, three different definitionshave been introduced from the consumerpoint of view: (i) quality is excellence; (ii)quality is value for money; and (iii) quality ismeeting or exceeding expectations. The firstdefinition displays some inherent weaknesses.For example, defining quality as being excel-lent is highly subjective and it varies from per-son to person. Although service quality isproposed as value for money, scholars arguedthat value and quality are two different con-structs (Bolton and Drew, 1991).

Defining quality as meeting or exceedingcustomer expectations is well established.Service quality is defined from the customerpoint of view and measured by the inferreddisconfirmation scale (best known as the ‘gapmodel’). Empirical studies, however, showedthat such a measurement causes validity andreliability problems (Teas, 1993). Recent liter-ature suggests that service quality is more rel-evant as to how well the service is delivered(Cronin and Taylor, 1992; Ekinci, 2002). Thisis also known as performance evaluation andis considered to influence customer satisfac-tion positively. The following fifth hypothesisis proposed to allude to this path.

H5: Service quality has a positive associationwith customer satisfaction.

Self-concept congruence

Two decades ago, Sirgy (1982) argued that con-sumers evaluate products by referring to theirself-concept. Self-concept and product imagesshare a degree of communality and, as such,there can be a degree of congruence betweenthe two. The idea is extended to suggest thatthe greater the degree of congruence, thehigher the probability of displaying specificbehaviour, such as intention to purchase or sat-isfaction. This theory has been applied in orderto examine the relationship between self-con-cept and different variables. Examplesincluded self-concept and preference forhouses (Malhotra, 1988), self-concept and storeimages (Sirgy and Samli, 1985), self-conceptand brand preferences, brand attitudes, pur-chase intentions (Hong and Zinkhan, 1995;Graeff, 1996), and self-concept and satisfactionwith holiday destinations (Chon, 1992).

Landon (1974) argued that the relation-ship between self-concept congruence andconsumer behaviour may differ across prod-uct categories due to involvement of differentself-concept (such as actual and ideal self).For example, the relationship between actualself-congruence and customer satisfactionmay not be significant because often con-sumers do not want to describe themselves,but to superimpose their ‘ideal’ self in pur-chase situations, particularly when the actualself-concept dimension is perceived to benegative. Later, Malhotra (1988) supportedthe idea of differential roles for actual, idealand social self-concept in product evaluation.His study suggested that ideal self-congru-ence rather than actual self-congruence hasthe primary influence on house preferences.Hamm and Cundiff (1969) reported a signifi-cant relationship between ideal self-congru-ence and product preference as opposed toactual self-congruence. More recently, Hongand Zinkhan (1995) showed that ideal self-congruence rather than actual self-concept isa better indicator of brand preference amongdifferent product categories such as cars andshampoos. Hence, not only the actual self butalso the ideal self-concept should be takeninto account when investigating the relation-ship between self-concept congruence andconsumer behaviour. Consequently, two typesof self-congruence are considered to be rele-vant to this study. The following hypotheses(H6 and H7) were developed to test thesepropositions.

H6: Actual self-concept congruence has a posi-tive association with customer satisfaction.

H7: Ideal self-concept congruence has a positiveassociation with customer satisfaction.

Desires congruence

The use of a comparison standard seems to becentral to the evaluation of both service qualityand customer satisfaction. Several comparisonstandards are introduced into the literaturefrom different perspectives such as expecta-tions, desires and experience-based norms.However, their utilization often triggeredmethodological problems in the measurementof service quality and customer satisfaction



due to their vague conceptualizations and mis-interpretation. Although customer expectationis the most frequently used one (Oliver, 1997;Parasuraman et al., 1988), the meaning ofexpectation is often mixed with desired out-comes. For example, Parasuraman et al. (1988)argued that the ‘should’ type of expectationmust be used to measure service quality as itreflects customers’ desires and wants. However,the empirical studies showed that this was nota good formulation as it caused various relia-bility and validity problems in measurement(Teas, 1993).

Although expectation is mixed withdesires in the service quality literature, thesetwo concepts are different. The latter is asso-ciated with consumer values. Employing val-ues (e.g. desires, wants) as a comparisonstandard is theoretically compelling becausethey are the centrepiece of human percep-tion and evaluation (Rokeach, 1973). Forexample, the means-end models imply thatproduct attributes are linked to consumer val-ues (Guttman, 1982). More recently, Ekinciand Chen (2002) showed that satisfactionwith hotel services differs between customerswho are divided into various segments by per-sonal values.

The early empirical studies reveal littlesupport for using values or desires as compar-ison standards (Westbrook and Reilly, 1983).One reason for the negative outcome isattributed to inadequate conceptualizationand poor measurement. Spreng et al. (1996)addressed the methodological issues experi-enced previously in value research. As aresult, they proposed a model by redefiningthe role of value, expectation, performanceand customer satisfaction. Their study indi-cates that the desires congruence that isdefined as the match or mismatch of what isdesired and actually received has a significantimpact on attribute satisfaction, informationsatisfaction and overall satisfaction.Consequently, the following hypothesis is pro-posed to evaluate such a stance.

H8: Desires congruence has a positive associa-tion with customer satisfaction.

Figures 16.2 and 16.3 illustrate two holisticmodels and the associated paths for the con-ceptual frameworks of this study.

Figure 16.2 illustrates that customer satis-faction is related to behavioural intentionsand attitudes. Also, the relationship betweensatisfaction and service quality is direct andfrom service quality to customer satisfaction.This model implies that attitudes, servicequality, actual, ideal and desires congruenceare antecedents of customer satisfaction. Asthe relationship between customer satisfac-tion and attitudes is bi-directional, attitudescan be a consequence of satisfaction.Together, an attitude towards service organi-zation and satisfaction stimulates customers’intention to visit and recommend behaviour.

H9: The satisfaction model significantly fits tothe data.


SQ

BI

ATT

IC

AC

CS

DC

H5

H6

H7

H8

H2

H3

H1

H4

CS

BI

ATT

IC

AC

SQ

DC

Fig. 16.2. Satisfaction model. SQ, service quality;AC, actual self-congruence; IC, ideal self-congruence; DC, desires congruence; CS, customersatisfaction; ATT, an attitude towards the serviceorganization; BI, behavioural intentions(recommend and return).

Fig. 16.3. Service quality model. For abbreviationssee Fig. 16.2.


The alternative model is also proposed byswapping the position of customer satisfac-tion with service quality. The followinghypothesis is relevant to this model.

H10: The service quality model significantlyfits to the data.

Methodology

Questionnaire development

The process of developing the questionnairerequires measurement and validation of thefollowing constructs: product concept, self-con-cept, attitude, desires congruence customer satisfac-tion, service quality and behavioural intention(recommend and return behaviour).

Measurement of actual self-congruence andideal self-congruence

Despite the fact that the theory of self-con-cept is compelling, empirical studies haveproduced mixed results. Some consumer psy-chologists argued that personality is a usefultool for understanding consumer behaviour,whereas others postulated that the use of per-sonality variables has negligible value. Forexample, Shank and Langmeyer (1993)reported a weak relationship between humanpersonality and brand image.

Although the aforementioned studies seemto oppose the self-congruence theory, a num-ber of methodological and theoretical short-comings contribute to these results. Amongthem is the inadequate conceptualization ofself-concept, poor instruments, weak method-ology, which fail to take into account the influ-ence of brand/product attributes, and themediate effect of personality variables(Malhotra, 1981, 1988). Moreover, a few stud-ies attempted to assess self-congruence usingstandard personality instruments that weredesigned with activities other than buying inmind. It should be noted that the attributes ofproduct concept could be very extensive anddifferent from the attributes of self-concept.Therefore, it may not be appropriate to defineself-concept by using the attributes of product

concept. To an extent, these considerationshave been taken into account in measuringself-concept congruence.

One of the recent debates involved inmeasuring self-concept congruence iswhether to use the gap score formula ordirect score formula (Sirgy and Su, 2000). Todate, the usual practice for measurement ofself-congruence has been to employ the gapscore formula. This measure indicates thedegree of match/mismatch between theproduct concept (e.g. restaurant, hotel, retailshop) and self-concept. To do this, theabsolute difference model was used to com-pute the self-congruence score (Ericsen andSirgy, 1992). Mathematically indicated;

n

ACk = S ÍPCik – ASCik Í (1)i = 1

where ACk = actual self-congruence score forrespondent (k); PCik = product concept scoreof respondent (k) along attribute (i); andASCik = actual self-concept score of respon-dent (k) along attribute (i).

One can note that the lower the score thehigher the actual self-congruence, since theabsolute difference model was employed. Thedirect score formula, on the other hand,requires neither self-ratings (actual or ideal)nor product ratings but measures the self-concept congruence on a numeric scale thatis facilitated by a scenario-type direction(Sirgy and Su, 2000).

The gap formula has received a number ofcriticisms. At the heart of them are inflatedreliability scores, spurious correlationsbetween theoretically related variables and amathematically computed gap score that maybe different from respondents’ actual evalua-tion (Peter et al., 1993). Despite these criti-cisms, the present study used this methodbecause of the need to make comparisonswith previous research. Furthermore, evi-dence that the direct formula is better thanthe gap formula is not very strong.

Malhotra (1981) recommended thatsemantic differential scales should be used tomeasure product and self-concept images.However, Landon (1974, p. 44), highlightedtwo issues regarding the use of this scalingprocedure. First, when measuring product



and self-concepts, the adjectives may corre-spond to different meanings and, therefore,research should ensure that both constructsare evaluated in the same direction and referto the same meaning. Second, as ratings ofactual and ideal self-concepts may beextremely sensitive to the social desirabilityeffect (Landon, 1974), those attributes thatare believed to suffer from this effect shouldbe eliminated. Armed with this knowledge, ascale was developed to measure both self- andproduct concepts as there was no genericscale available for the evaluation of services.

The scale development procedure involveda number of testing stages. Firstly, 58 personal-ity traits were elicited from the literature onthe basis that they described both people andproducts (Malhotra, 1981; Graeff, 1996; Aaker,1997). Secondly, the content of these items waschecked to ensure that the selected adjectiveswould be relevant to describe a restaurant. Tothis end, a pilot study used a small group ofBritish subjects (n = 26, 48% male, 52% female)from a wide spectrum of age groups (16 to55). The criterion for selection of an adjectivewas if it was chosen by 70% of the sample. Thisresulted in 12 pairs of adjectives: exciting/dull,organized/disorganized, formal/informal,popular/unpopular, extravagant/economical,modern/classical, sophisticated/unsophisti-cated, friendly/unfriendly, clean/dirty, com-fortable/uncomfortable, pleasant/unpleasantand business oriented/family oriented.

Thirdly, the above adjectives were tested todetermine their applicability to both peopleand products (i.e. a restaurant). This involvedassessing the polarity of the adjectives and test-ing for the social desirability effect and wasaccomplished by a content analysis. Twentysubjects (50% male, 50% female) completed aquestionnaire containing the pairs of adjectivesqualified earlier. Subjects were then inter-viewed by the researchers about their ratings.The attributes were then judged based onthree criteria. Firstly, the subjects needed tofeel comfortable using the adjectives in bothcontexts; secondly, the meaning of both appli-cations should have been the same; and thirdly,there had to be no interference from the socialdesirability effect (Landon, 1974). As a result,three of the 12 items were deleted. These wereclean/dirty, comfortable/uncomfortable and

pleasant/unpleasant. Eight pair of adjectivesqualified from this selection process:exciting/dull, organized/disorganized, formal/infor-mal, popular/unpopular, extravagant/economical,modern/classical, sophisticated/unsophisticated andfriendly/unfriendly.

The product concept was measured usinga seven-point (�3 to +3) numeric scale.Actual self-concept was measured using thesame scale but with the numeric points of thescale changed to 1–7 to reduce the halo effect(Sirgy, 1982). The following direction wasgiven to measure actual self-concept.

We would like you to describe yourself as you actuallyare. First, think about how you see yourself. Pleasedescribe some characteristics of your personality using thefollowing scales (e.g. friendly, organized) below. Mark(X) the number that best represents how you see yourself.

Ideal self-concept was operationalized on thesame scale by using the following instruction.

This time, we would like you to describe your ideal per-sonality. Think about the type of person that you wouldideally like to be. Please go back to the same scale aboveand CIRCLE the number that represents how youwould ideally like to see yourself. Do not worry if youractual self-rating and ideal self-rating coincide.

Measurement of remaining constructs

Satisfaction with services was assessed by twoseven-point numeric scales. The labels forthese scales were worse than my expectations/bet-ter than my expectations and completely dissatis-fied/completely satisfied (Spreng and Mackoy,1996). The customers’ attitude towards therestaurant was measured by a seven-pointnumeric scale. The scale items were: bad/good,valuable/worthless, nice/awful, positive/negativeand dislike/like (Maio and Olson, 1994).

Evaluation of overall service quality wasmeasured using a seven-point numeric scalewith (1) being extremely low quality and (7)being extremely high quality. Desires congru-ence was measured by two-item scale devel-oped by Spreng and Mackoy (1996). Finally,the customers’ behavioural intentions (rec-ommend and return) were measured by twoseven-point numeric scales with (1) repre-senting extremely unlikely and (7) extremely likely(Cronin and Taylor, 1992).



Application of the questionnaire

The study took place in a university campusenvironment due to sampling convenience.The campus contained eight restaurants andfrom these the one that offered a modern ser-vice style with different types of food anddrink throughout the day was chosen. A ran-dom sample of campus occupants was soughtand, to this end, 500 questionnaires were sentout to British nationals through the universityinternal mail.

At the end of the 5-week period, a total of109 usable questionnaires was returned(22%). The sample was 67% female, 33%male. Forty-nine per cent of the respondentswere between 16 and 24, 25% between 25 and34, 26% were 35 years of age or above. Forty-three per cent of the sample was students and57% was staff. The majority of respondents(65%) made more than four visits to therestaurants. Thirty per cent made two tothree visits and 5% made only one visit. Thevisits were on different occasions and at dif-ferent times of the day but were mostlyaround lunchtime (59%).

Findings

The principal objective of this study was totest the two competing models that outlinethe relationship between customer satisfac-tion and other variables. Prior testing of themodels, validity and reliability of the mea-sures were established.

Validity and reliability of measurements

The first stage of analysis involved testing thedimensionality of the product and self-con-cept scales. To this end, three separateexploratory factor analyses were conductedfor the product concept, actual self-conceptand ideal self-concept scales using principalcomponent extraction with Varimax rotation(Hair et al., 1998). Initial findings suggestedthat the product concept scale consisted oftwo dimensions, whereas the self-conceptscales consisted of three dimensions.Interestingly, the first factor was identicalacross the three factor analyses. This factorcontained the following items: (i)exciting/dull, (ii) organized/disorganized,(iii) sophisticated/unsophisticated, (iv) popu-lar/unpopular and (v) friendly/unfriendly.The first factor was retained for two reasons;firstly, it explained most of the variance in theanalyses and secondly, the reliability of theother factors was unacceptable (all alphacoefficient values were < 0.50).

Table 16.1 shows the outcome of the fac-tor analysis with Varimax rotation for theproduct concept scale.

Items of the product concept scale wereloaded on the same factor. The level of vari-ance explained by this solution was low butacceptable (54%) and this finding providedevidence for the convergent validity of themeasure (Hair et al., 1998).

Table 16.2 shows the outcome of the fac-tor analysis with Varimax rotation for theactual self-concept scale.


Table 16.1. The product concept scale: factor analysis with Varimaxrotation.

Factor loadinga

The product concept scale Factor 1 Communality

Dull/exciting 79 62Disorganized/organized 76 58Unpopular/popular 61 38Unsophisticated/sophisticated 77 59Unfriendly/friendly 72 53

Eigenvalue 2.72Explained variance 54.43%

aNumbers are magnitudes of the factor loading multiplied by 100.


According to Table 16.2, these results weresimilar to the product concept scale and pro-vided evidence of convergent validity of theactual self-concept scale (Hair et al., 1998).The ideal self-concept scale also producedsimilar results by extracting 53% of variancein the data set.

Item-to-total correlation coefficients forthe restaurant concept scale ranged from0.44 to 0.64 and the actual self-concept scaleranged from 0.39 to 0.61. The reliabilityscores of the two scales exceeded the mini-mum recommended internal consistencythreshold (alpha coefficient ≥0.70) and there-fore the scores estimated by these scales canbe considered as reliable (Churchill, 1979).The item-to-total correlation score for theideal self-concept scale ranged from 0.39 to0.53 and the reliability of this scale was alsoacceptable. The reliability of the attitudescale (alpha coefficient = 0.87) was excellent.The item-to-total correlation for this scaleranged from 0.60 to 0.78 and thus there wasno need to eliminate any item from the scale.From the internal consistency reliability mea-sure, the customer satisfaction (0.86) andbehavioural intention scales (0.90) were alsodeemed to be reliable.

Testing of models

The structural models were tested usingMaximum Likelihood estimator of LISREL-VIIIcausal modelling procedure (Jöreskog andSörbom, 1996). This testing determined the

magnitude of individual relationships, themodels’ goodness of fit, and the hypothesizedpaths. PRELIS was used to generate the vari-ance–covariance matrix as input.

The overall fit of the structural model wasdetermined initially by examining the chi-squared statistics for each model. A signifi-cant chi-squared statistic indicates aninadequate fit but this statistic is sensitive tosample size and model complexity. Thereforerejection of a model on the basis of this evi-dence alone is inappropriate (Hair et al.,1998). Other measures of fit compensatingfor sample size were also applied. These aregoodness of fit index (GFI), adjusted good-ness of fit index (AGFI), normed fit index(NFI), comparative fit index (CFI) and rootmean square error of approximation(RMSEA). Figure 16.4 shows testing of thesatisfaction model and its findings.

As can be seen from the chi-squared statis-tics and the associated probability value (P >0.05, not significant), the data fit the satisfac-tion model (chi-squared for the researchmodel was 9.17 with seven degrees of free-dom). The other fit indices also showed thatthe model has a good fit as these estimatesare well above the recommended thresholds(Hu and Bentler, 1999). The model alsoexplained a relatively high proportion of thevariation in behavioural intention (60%).The path model explained 93% of the vari-ance in predicting customer satisfaction and35% of variance in estimating attitudestowards restaurant. The service quality modelwas tested by using the same procedure but


Table 16.2. The actual self-concept scale: factor analysis with Varimaxrotation.

Factor loadinga

The actual self-concept scale Factor 1 Communality

Dull/exciting 74 56Disorganized/organized 56 31Unpopular/popular 70 49Unsophisticated/sophisticated 79 63Unfriendly/friendly 75 56

Eigenvalue 2.57Explained variance 51.4%

aNumbers are magnitudes of the factor loading multiplied by 100.


the data did not fit to this model (chi-squared= 7: 25.87, P = 0.000, GFI = 0.94, CFI = 0.93,AGFI = 0.76, NFI = 0.92, RMSEA = 0.16).Therefore, hypothesis 10 was rejectedwhereas hypothesis 9 was accepted.

Given this support, the standardized esti-mates for the model paths and the associatedt-values for the satisfaction model are pro-vided in Fig. 16.4. Of the eight proposedrelationships, seven were statistically signifi-cant. For the path leading to behaviouralintention, from both customer satisfaction (t-value = 4.49) and attitudes (t-value = 5.51)are positive and highly significant. Theseresults fully supported hypotheses 1 and 3.Beta values indicated that attitudes havehigher impact (b = 0.34) on the intention torecommend/return behaviour than cus-tomer satisfaction (b = 0.11).

Hypotheses 2 and 3 that testing themutual relationship between customer satis-faction and attitudes were supported, as thetwo paths were statistically significant (t-values= 8.07 and 2.11). As can be seen from thegamma values, the effect of attitudes on satis-faction (2.83) was positive and higher thanthe effect of satisfaction on attitudes (0.39).Hypotheses 5, 7 and 8 were fully supported asthe paths from service quality (t-value = 3.38),ideal self-congruence (t-value = 2.73) anddesires congruence (t-value = 2.25) to cus-tomer satisfaction were statistically significant.

However, hypothesis 6 was rejected as therelationship between actual self-congruenceand satisfaction was not statistically significantat the 0.05 alpha level (t-value = 0.06).

Conclusion

Despite its exploratory nature, the findingssuggest that service quality is an antecedentof customer satisfaction and therefore evalua-tion of service quality leads to customer satis-faction. However, there was no evidence tosupport the opposite relationship. This find-ing contributes further to the debate regard-ing the difference between customersatisfaction and service quality (Ekinci andRiley, 1998; Fournier and Mick, 1999). Thestudy also indicates that customer satisfactionrather than service quality is a better reflec-tion of overall attitudes, as suggested byOliver (1980). As the relationship betweencustomer satisfaction and attitudes are recip-rocal, an attitude not only serves as anantecedent of customer satisfaction but also aconsequence. Furthermore, these two con-cepts have significant impact on customers’behavioural intentions. By the same token,both customer satisfaction and attitudesshould be taken into account in predictingcustomers’ intention to recommend andreturn behaviour.


Service Q

Actual SC

Ideal SC

Desires C

CustomersatisfactionR2 = 93%

AttitudesR2 = 35%

BIR2 = 60%

5.31 (3.38)*

0.02 (0.6)

1.17 (2.73)*

0.23 (2.25)*

0.39 (8.07)* 2.83 (2.11)*

34 (5.51)*

0.11 (4.49 )*

Fig. 16.4. Satisfaction model: antecedents and consequences of satisfaction. *Significant P <0.05,**significant P <0.01. Model fit statistics chi-squared = (7: 9,17, P = 0.24, not significant), GFI = 0.98, CFI =0.99, AGFI = 0.91, NFI = 0.97, RMSEA = 0.05.


The study offers evidence for the involve-ment of self-concept in the evaluation of ser-vices due to the fact that the ideal self-conceptwas found to be an antecedent of customer sat-isfaction. This supports the observations ofMalhotra (1988) and Landon (1974) that therole of the self-concept varies and that eitheractual or ideal self can be the dominant char-acter in different situations. Exactly how thisvariation occurs depends on the situation.According to Graeff (1996, p. 16), for exam-ple, the ideal-self concept may be more rele-vant to publicly consumed as opposed toprivately consumed products. Hong andZinkhan (1995) noted that such a result isprobably attributed to consumers’ strongdesire to reach their ideal state, which willserve to improve their self-esteem. One pos-sible explanation may be that when there areseveral restaurants available, a specific restau-rant may upgrade consumers’ actual self toideal self-concept. The restaurant used in thisstudy differentiates itself from the other avail-able restaurants on campus in terms of offer-ing new concepts with brand new facilities.Thus, it aims to satisfy consumers’ higherneeds (e.g. friendliness, attractive environ-ment) as well as functional needs (e.g. conve-nient place for eating). This might help toexplain why the evaluation was strongly relatedto ideal self rather than actual self in this study.

An additional contribution of this study isthe investigation of the relationship betweendesires congruence and customer satisfac-tion. The results indicate that customers usetheir desires as a comparison standard intheir satisfaction decision. This is in line withthe notion that values should be seen as oneof the antecedents of customer satisfaction(Spreng et al., 1996).

In pragmatic terms, the analysis suggestedthat service quality, personality of customers,

their desires and attitude towards the serviceorganizations should be taken into accountin measuring customer satisfaction. By impli-cation, the survey questionnaires should con-tain these questions in order to draw a truepicture of customer satisfaction. Also, thefindings relating to the ideal self-conceptcongruence and desires congruence mayhelp managers to improve marketing com-munications. For example, advertising mes-sages should contain the desirablepersonality traits (e.g. friendliness, excite-ment) in order to develop a positive attitudetowards the service organization.Alternatively, these traits could help man-agers to position service organizations incompetitive markets. The whole idea of self-congruence measures implies that managersshould take into account personality of theircustomers in developing better products.The higher the self-concept congruencemeans the higher the satisfaction. By implica-tion, service providers should customize deliv-ery of services according to customers’personality traits. For example, if a customerwas identified as being egocentric, the strategyof delivering services would be different froma traditional customer.

Although, the study makes important theo-retical contributions to the understanding ofthe antecedents and consequences of cus-tomer satisfaction, it nevertheless entails cer-tain limitations, which have to be taken intoaccount when interpreting the findings. Oneof the limitations of the study is the use ofnon-probability sampling (convenience sam-ple) to validate the underlying theory. Thefindings are specific to one culture (Britishnationals) and one service organization(restaurants). Also, the sample size is smalland therefore the findings cannot be general-ized to the whole population.


References

Aaker, J.L. (1997) Dimensions of brand personality. Journal of Marketing Research 34, 347–356.Ajzen, I. and Fishbein, M. (1980) Understanding Attitudes and Predicting Social Behavior. Prentice-Hall,

Englewood Cliffs, New Jersey.Allport, G. (1935) Attitudes. In: Murchinson, C.A. (ed.) A Handbook of Social Psychology. Clark University

Press, Worcester.Bolton, R.N. and Drew, J.H. (1991) A multistage model of customers’ assessment of service and value.

Journal of Consumer Research 17, 375–384.


Burnkrant, R.E. and Page, J.P., Jr (1982) An examination of the convergent, discriminant, and predictivevalidity of Fisbein’s behavioral intention model. Journal of Marketing Research 19, 550–561.

Chon, K. (1992) Self-image/destination image congruity. Annals of Tourism Research 19, 360–362.Churchill, G.A. (1979) A paradigm for developing better measure of marketing constructs. Journal of

Marketing Research 16, 64–73.Churchill, G.A. Jr and Suprenanat, C. (1982) An investigation into the determinants of customer satisfac-

tion. Journal of Marketing Research 19, 491–504.Cronin, J.J., Jr and Taylor, S.A. (1992) SERVPERF versus SERVQUAL: reconciling performance-based and

perception-minus-expectations measurement of service quality. Journal of Marketing 58, 15–131.Ekinci, Y. (2002) A review of theoretical debates on the measurement of service quality: implications for

hospitality research. Journal of Hospitality and Tourism Research 26, 199–216.Ekinci, Y. and Chen, J. (2002) Segmenting oversees British holidaymakers by personal values. Journal of

Hospitality and Leisure Marketing 9, 1–5.Ekinci, Y. and Riley, M. (1998) A critique of the issues and theoretical assumptions in service quality mea-

surement in the lodging industry: time to move the goal posts? International Journal of HospitalityManagement 17, 349–362.

Ericsen, M.K. and Sirgy, M.J. (1992) Employed females’ clothing preference, self-image congruence, andcareer anchorage. Journal of Applied Social Psychology 22, 408–422.

Fournier, S. and Mick, D.G. (1999) Rediscovering satisfaction. Journal of Marketing 63, 5–23.Graeff, T.R. (1996) Using promotional messages to manage the effects of brand and self-image on brand

evaluations. Journal of Consumer Marketing 13, 4–18.Guttman, J. (1982) A means-end chain model based on consumer categorisation process. Journal of

Marketing 46, 60–72.Hair, J.F., Jr, Anderson, R.E., Tatham, R.L. and Black, W.C. (1998) Multivariate Data Analysis. Prentice-Hall,

Englewood Cliffs, New Jersey. Hamm, B.C. and Cundiff, E.W. (1969) Self-actualisation and product perception. Journal of Marketing

Research 6, 470–472.Hong, J.W. and Zinkhan, G.N.M. (1995) Self-concept and advertising effectiveness: the influence of con-

gruence, conspicuousness and response mode. Psychology and Marketing 12, 53–77.Hu, L. and Bentler, P.M. (1999) Cut-off criteria for fit indexes in covariance structure analysis: conven-

tional criteria versus new alternatives. Structural Equation Modelling 6, 1–55Jöreskog, K.G. and Sörbom, D. (1996) LISREL 8 User’s Reference Guide. Scientific Software International,

Chicago, Illinois.Katz, D. (1960) The functional approach to attitudes. Public Opinion Quarterly 24, 163–204. Landon, E.L. (1974) Self-concept, ideal self-concept, and consumer purchase intentions. Journal of

Consumer Research 1, 44–51.LaTour, S.A. and Peat, N.C. (1979) Conceptual and methodological issues in consumer satisfaction

research, In: William, L.W. (ed.) Advances in Consumer Research. Association for Consumer Research,Ann Arbor, Michigan, pp. 431–437.

Maio, G.R. and Olson, J.M. (1994) Value-attitude-behavior relations: the moderating role of attitude func-tions. British Journal of Social Psychology 33, 301–312.

Malhotra, N.K. (1981) A scale to measure self-concepts, person concepts and product concepts. Journal ofMarketing Research 18, 456–464.

Malhotra, N.K. (1988) Self concept and product choice: an integrated perspective. Journal of EconomicPsychology 9, 1–28.

Mittal, V., Ross, W.T. Jr and Baldasare, P.M. (1998) The asymmetric impact of negative and positiveattribute-level performance on overall satisfaction and repurchase intentions. Journal of Marketing 62,33–47.

Oh, H. and Parks, S. (1997) Evaluating the role of attribute importance as a multiplicative weighting vari-able in the study of hospitality consumer decision-making. Journal of Hospitality and Tourism Research21, 62–80.

Oliver, R. (1980) A cognitive model of the antecedents and consequences of satisfaction decision. Journal ofMarketing Research 17, 460–469.

Oliver, R.L. (1997) Satisfaction: a Behavioral Perspective on the Consumer. McGraw-Hill Company, London.Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1988) SERVQUAL a multiple-item scale for measuring

consumer perception of service quality. Journal of Retailing 64, 13–40.



Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1994) Alternative scales for measuring service quality: acomparative assessment based on psychometric and diagnostic criteria. Journal of Retailing 70,193–199.

Peter, J.P., Churchill, G.A. and Brown, T.J. (1993) Cautions in the use of difference scores in consumerresearch. Journal of Consumer Research 19, 655–662.

Peterson, R.A. and Wilson, W.R. (1992) Measuring customer satisfaction: fact or artefact. Journal of theAcademy of Marketing Science 20, 58–66.

Reeves, C.A. and Bednar, D.A. (1994) Defining quality: alternatives and implications. Academy ofManagement Review 19, 419–445.

Rokeach, M. (1973) The Nature of Human Values. The Free Press, New York.Shank, M.D. and Langmeyer, L. (1993) Does personality influence brand image. The Journal of Psychology

128, 157–164.Sirgy, M.J. (1982) Self-concept in consumer behavior: a critical review. Journal of Consumer Research 9,

287–300.Sirgy, M.J. and Samli, A.C. (1985) A path analytic model of store loyalty involving self-concept, store image,

geographic loyalty, and socio-economic status. Journal of the Academy of Marketing Science 13, 265–291. Sirgy, M.J. and Su, C. (2000) Destination image, self-congruity, and travel behavior: toward an integrative

model. Journal of Travel Research 38, 340–352.Spreng, R.A. and Mackoy, R.D. (1996) An empirical examination of a model of perceived service quality

and satisfaction. Journal of Retailing 72, 201–214.Spreng, R.A., MacKenzie, S.B. and Oshavsky, R.W. (1996) A re-examination of the determinants of con-

sumer satisfaction. Journal of Marketing 60, 15–32Teas, R.K. (1993) Expectations, performance evaluation, and consumers’ perceptions of quality. Journal of

Marketing 57, 18–34.Westbrook, R.A. and Reilly, M.D. (1983) Value-precept disparity: an alternative to the disconfirmation of

expectations theory of consumer satisfaction, In: Bagozzi, R.P. and Tybout, A.M. (eds) Advances inConsumer Research. Association for Consumer Research, Ann Arbor, Michigan, pp. 256–261.



Chapter seventeenFirst-time and Repeat Visitors to Orlando, Florida: aComparative Analysis of Destination Satisfaction

Paul Fallon and Peter SchofieldSchool of Leisure, Hospitality and Food Management, University of Salford, Frederick Road,Salford M6 6PU, UK

Abstract

The chapter compares first-time and repeat visitor satisfaction with Orlando, Florida. Factor analysis (PCA) ofsubjects’ ratings on 22 ‘performance’ attributes produced five factors: ‘primary’, ‘secondary’ and ‘tertiary’attractions, ‘facilitators’ and ‘transport plus’. A one-way between-groups multivariate analysis of variance iden-tified a significant difference between first-time and repeat segments on the ‘secondary’ attractions; regres-sion of overall tourist satisfaction with Orlando against the factors showed that ‘secondary’ attractions werethe single most influential factor affecting tourists’ overall satisfaction with Orlando. Subdivision of the sam-ple into first-timer and repeater segments showed that the overall satisfaction of first-timers and repeaters wasexplained by different ‘hierarchies’ of factors. First-timers’ overall satisfaction was explained by a four-factormodel with ‘facilitators’ accounting for the dominant contribution and ‘secondary’ and ‘primary’ attractionsalso having significant influence. By comparison, repeater satisfaction was explained by a five-factor modelwith ‘secondary’ attractions carrying the most weight followed by ‘primary’ attractions and ‘facilitators’.

Introduction

Customer satisfaction has been defined aspost-consumption evaluative judgement(Oliver, 1980; Churchill and Surprenant, 1982;Westbrook and Oliver, 1991) that representsthe ‘outcome’ for the customer after exposureto the service product (Baker and Crompton,2000; Kozak, 2001). By comparison, qualityrefers to the service operation’s ‘output’, i.e.the attributes of the product that are primarilyunder the control of the operation (Cromptonand Love, 1995; Schofield and Fallon, 2000).Nevertheless, it should be emphasized that sat-

isfaction also represents a potentially signifi-cant ‘outcome’ for the operation’s ‘output’ interms of internal benefits – such as resourceanalysis, product enhancement and differenti-ation – and external benefits – such as cus-tomer loyalty and positive word-of-mouthrecommendation. Given these benefits, it is nosurprise that the measurement of tourist satis-faction has become a major area of research inthe last three decades (Kozak, 2001).Furthermore, due to the key role played bydestinations in most holidays, the measure-ment of tourist satisfaction at this level wouldseem particularly relevant.



First-time and Repeat Visitors

The satisfaction measurement debate is fur-ther complicated by the influence of previousexperience (Crompton and Love, 1995;Kozak, 2001). Research comparing first-timeand repeat visitor behaviour at destinationshas been the subject of an increasing amountof research in recent years. A number of stud-ies have identified significant differencesbetween first-time and repeat visitors totourist destinations with respect to behaviourand experience. For example, first-time visi-tors are more likely to seek variety and newexperiences, whereas repeaters will tend tochoose familiar places (Gyte and Phelps,1989; Mazursky, 1989; Watson et al., 1991;Gitelson and Crompton, 1994). Repeaters’behaviour may reflect a variety of motivesincluding: risk reduction; emotional attach-ments; a desire to show the destination toother people; and also the fact that repeatersare more likely to be seeking relaxation thanfirst-timers. However, variability in behaviourpatterns within this group is also likely, forexample some repeaters may be more activethan others because they wish to explore thedestination further (Gitelson and Crompton,1984). More recently, Oppermann (1997)found significant differences between thebehaviour of first-time and repeat visitors toNew Zealand. First-timers appeared to bemuch more active than repeaters, in that theyvisited many more attractions and sites in thedestination area; interestingly, this includedboth popular and lesser-known sites. By com-parison, repeaters visited considerably lessattractions and destinations despite theirlonger stay, indicating that their impact ismore geographically concentrated on fewerlocations and attractions than that of first-timers. This suggests that certain locationsthat are not perceived as attractive enoughwill not be selected for repeat visits(Oppermann, 1996).

Despite the importance of the repeat visi-tor segment for many attractions and desti-nations, especially mature destinations(Kozak, 2001), and the increasing attentionbeing paid to repeaters in empiricalresearch, the factors of significance in

repeater destination satisfaction have beenneglected. The problem is of theoreticalinterest and the results have practical mar-keting applications with respect to destina-tion enhancement and promotion. Thisresearch attempts to address the weaknessesidentified in previous tourist satisfactionresearch by using a refined methodology,specifically taking into account the role ofprevious visitation on tourist satisfaction witha destination – Orlando, Florida. The desti-nation-level and prior visitation aspects ofthe study have been highlighted as impor-tant new dimensions of tourist satisfactionmeasurement (Kozak, 2001).

Methodology

Instrumentation

Empirical comparisons of the reliability andvalidity of alternative satisfaction modelsbased on visitation to camp sites (Dorfman,1979; Fick and Brent Ritchie, 1991), events(Crompton and Love, 1995) and restaurants(Yuksel and Rimmington, 1998) have sup-ported the case for a single measurementbased on performance, notwithstanding thecomplex nature of satisfaction due to theinfluence of a wide range of personal and sit-uational variables, such as needs, disposition,expectations, nationality and travelling com-panions (Ekinci et al., 2000; Kozak, 2001).Tourists, even first-timers, become moreexperienced over the course of their holidaydue to its longitudinal nature, and conse-quently have the potential to refine their ini-tial expectations (Danaher and Mattsson,1994; Weber, 1997). From this perspective,the performance-only conceptualization ofsatisfaction would seem to be a more theoret-ically valid approach than one based on a(dis)confirmation model. Indeed, the perfor-mance only paradigm is now widely regardedas the most effective construct for satisfactionmeasurement (Churchill and Surprenant,1982; Carman, 1990; Cronin and Taylor,1992). Meyer and Westerbarkey (1996) arguethat measurements that focus on perceptionsof performance alone are more typical of thecognitive process, and Yuksel and

204 P. Fallon and P. Schofield


Rimmington (1998, p. 63) propose that ‘per-formance bears a pre-eminent role in theformation of customer satisfaction because itis the main feature of the consumptionexperience’. The ‘performance-only’ con-struct was therefore adopted in this study toexamine first-time and repeat visitor satisfac-tion with Orlando.

The attributes on which Orlando was eval-uated were generated from the triangulationof primary and secondary methods (Jenkins,1999; Tribe and Snaith, 1999; Oh, 2001).Secondary research took the form of a reviewof both the relevant academic and commer-cial literature, including research papers ondestination image, quality and satisfaction,and brochures and travel guides respectively.Preliminary primary research incorporatedfree elicitation during eight focus groups andan open-ended questionnaire distributed to astratified random sample of employees at theUniversity of Salford. In both cases, subjectswere representative of Orlando’s UK market.There was consensus on a relatively parsimo-nious set of elements on which UK visitorsmake judgements on Orlando and a distinc-tion between the destination’s offering of spe-cific attractions and activities, which weredominated by its primary attractions such astheme parks, and generic facilities needed toenjoy these attractions during the holiday,such as accommodation. This procedure pro-duced 22 attributes that were incorporatedinto a performance-only construct within aquestionnaire survey.

The questionnaire required respondentsto rate Orlando’s attributes according totheir performance levels on their currentholiday. The performance scale anchorswere ‘extremely poor’ (1) and ‘extremelygood’ (7) with all intervening options clearlylabelled. Data relating to personal details,overall satisfaction and visitor intention toreturn to Orlando and recommend the des-tination to others were also collected.Overall satisfaction and intention to returnand recommend Orlando were measured onseven-point Likert-type scales. The inclusionof these three measures also facilitated ananalysis of the performance scale’s reliabilityand construct validity (Yuksel andRimmington, 1998).

The sample

After an initial pilot study, which resulted inonly minor amendments, a post-visit conve-nience sample of 467 UK visitors to Orlandowas taken at Manchester (UK) and OrlandoSanford (USA) airports in September 2001.Orlando was chosen as the destination sub-ject primarily because it is the UK’s mostpopular long-haul holiday destination with1.31 million UK visitors in 2000 – 43.5% ofoverseas visitors to Orlando (Orlando CVBResearch, 2001). At Manchester airport,subjects were intercepted after checking in,en route to the departure lounge; atSanford airport, subjects were approachedin the departure lounge. The Manchestersurvey produced 141 usable questionnairesand the Sanford sample produced 326.There were no significant differences (P >0.05) between the samples on the post-visitratings of Orlando’s attributes. On thisbasis, they were merged; use of such a multi-ple sample has been proposed by a numberof authors to compensate for the practicalproblems encountered in similar surveys(Oliver, 1997; Yuksel and Rimmington,1998). First-timers were outnumbered byrepeaters in both the Manchester(30%:70%) and Sanford (35%:65%) sur-veys. This reflects the generally high level,i.e. 72%, of repeat visitation to Orlando(Orlando CVB Research, 2001) and tomature destinations in general (Kozak,2001). The majority (90%) of tourists in theoverall sample stayed in Orlando for 2weeks, which, even given the scale ofOrlando’s offering, gave them a reasonabletime to familiarize themselves with the desti-nation. Most (70%) were travelling in par-ties of four or more; these were mainlyfamily groups.

Data analysis

The data were analysed using SPSS Version11. A factor analysis, using principal compo-nents as the method of extraction, withVarimax rotation was conducted on the sub-jects’ ratings on each of the 22 variables toreduce multi-collinearity and identify a

First-time and Repeat Visitors to Orlando, Florida 205


smaller set of factors with eigenvaluesgreater than or equal to 1.0 and factor load-ings greater than 0.4 (Stevens, 1992).Cronbach’s alpha coefficient, a Kaiser–Meyer–Olkin (KMO) test of sampling ade-quacy and Bartlett’s test of sphericity werecomputed to determine the factorability ofthe correlation matrix. A one-way between-groups multivariate analysis of variance(MANOVA) was performed, after prelimi-nary assumption testing, to investigate thedifferences between first-time and repeatersegments on the factor scores. Finally, multi-ple regression analysis was employed toexamine the factors of significance in first-time and repeat visitor satisfaction withOrlando.

Results and Discussion

Factor analysis of Orlando’s attributeperformance ratings

The analysis produced a five-factor solution(with eigenvalues >1.0) which explained56.53% of the overall variance before rota-tion; 15 of the 21 items had loadings greaterthan 6.0, indicating a good correlationbetween the items and the factor groupingsthey belong to. The KMO value of 0.878 was‘meritorious’ (Kaiser, 1974) and the Bartlett’stest of sphericity reached statistical signifi-cance, supporting the factorability of the cor-relation matrix. The results, given in Table17.1, seem to support the findings from the


Table 17.1. Results of the factor analysis of Orlando’s attribute performance ratings.

Orlando’s attributes Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Communality

Factor 1: FacilitatorsAccommodation 0.803 0.685

Cleanliness 0.763 0.660Pool 0.744 0.592Safety 0.719 0.598Customer service 0.718 0.626Friendliness of locals 0.513 0.465

Factor 2: Secondary attractionsGoods at bargain prices 0.802 0.679Shopping facilities 0.770 0.665Restaurant VFM 0.753 0.670Variety of restaurants 0.666 0.534 Opportunity for rest and relaxation 0.457 0.489

Factor 3: Tertiary attractionsNatural and wildlife attractions and trails 0.780 0.629

Cultural and historic attractions and trails 0.775 0.639

Sports facilities 0.621 0.477Bus service 0.514 0.338Nightlife 0.440 0.438

Factor 4: Core attractionsMany things to see and do 0.808 0.736Something for everyone 0.743 0.673Theme parks 0.675 0.535

Factor 5: TransportCar-hire service 0.811 0.696Road signs that are easy to follow 0.401 0.388

Eigenvalue 6.244 2.007 1.686 1.434 1.067Variance (%) 28.381 9.123 7.664 6.520 4.851Cumulative variance (%) 28.381 37.504 45.168 51.688 56.539Cronbach’s alpha 0.8502 0.7896 0.7888 0.7382 0.5224Number of items (total = 21) 6 5 5 3 2


qualitative research at the front end of thestudy in terms of the distinction which wasmade between Orlando’s attractions (e.g. itstheme parks) and its secondary elements (e.g.accommodation and customer service), whichfacilitated the enjoyment of the main features.Indeed, there appears to be a good fitbetween the factors and Kotler et al.’s (1999)‘product level concept’ in that core, secondaryand tertiary attractions, facilitators and transportplus were identified. The core, secondary andtertiary attractions represent the ‘pull’ ele-ments, whilst the facilitators and transportplus groupings enable the attractions to beexperienced and optimized by the tourist.

A one-way between-groups MANOVA wasused to investigate the differences betweenfirst-time and repeat visitors on the factorscores. Preliminary assumption testing usingLevene’s test of equality of error variances (P >0.05) and Box’s test of equality of vari-ance–covariance matrices (P = 0.01) showedno significant violations. There was a statisti-cally significant difference between first-timeand repeat visitors on the combined factors: F = 4.39, P <0.01; Wilks’ Lambda = 0.96; par-tial eta squared = 0.04. When the results forthe factors were considered separately, theonly difference to reach statistical signifi-cance using a Bonferoni adjusted alpha levelof 0.01 (Tabachnick and Fidell, 1996) was onFactor 2 (secondary attractions): F = 16.96, P <0.01; partial eta squared = 0.04; only 4% ofthe variance in Factor 2 is explained by first-time or repeat visitor status.

The regression of ‘overall satisfaction withOrlando’ against the five factors showed thatFactor 2 was the single most influential factoraffecting tourists’ overall satisfaction with thedestination; a 1-unit increase in the perfor-mance of the secondary attractions wouldlead to a 0.330-unit increase in tourists’ overalllevel of satisfaction, all other variables beingheld constant. Additionally, the performanceof 11 of the 22 Orlando attributes were ratedsignificantly higher by repeat visitors thanfirst-timers (P <0.05). Consequently, the sam-ple was subdivided into first-time visitors andrepeaters to analyse both the attribute load-ings on the factors associated with each seg-ment and the variance in each segment’soverall satisfaction explained by the factors.

Factor analysis of first-time and repeat visitorratings

The results of the factor analysis of first-timer and repeater satisfaction ratings onOrlando’s attributes are given in Tables 17.2and 17.3. In both cases, the analysis pro-duced a five-factor solution (with eigenval-ues >1.0) which explained 58.78%(first-timers) and 56.80% (repeaters) of theoverall variance before rotation; for bothsegments, 16 of the 21 items had loadingsgreater than 6.0, indicating a good correla-tion between the items and the factor group-ings they belong to. The KMO values of0.841 (first-timers) and 0.856 (repeaters)were ‘meritorious’ (Kaiser, 1974) and theBartlett’s test of sphericity reached statisticalsignificance, supporting the factorability ofthe correlation matrix.

Factor 1 – facilitators – loads on genericand functional attributes that are notenough in themselves to attract visitors, buttheir presence enables and supplementsenjoyment of the destination and its attrac-tions. Furthermore, as such they offer aframe of reference for comparison of onedestination with another. Interestingly,‘friendliness of the locals’ loads on this fac-tor for first-timers only, whereas in the caseof repeaters, this attribute loads on to thetransport plus factor, which also representsfacilitating features.

Factor 2 is comprised of secondary attrac-tions. Whilst Orlando’s theme parks remainits primary attraction, the destination isbecoming increasingly well known for itsshopping and eating facilities, which washighlighted in both the secondary researchand the results from the open-ended ques-tions and focus groups. For example,Orlando CVB Research (2001) identifiedshopping and dining in restaurants as thetop two holiday activities, outstripping visit-ing the theme parks, for UK visitors in 2000.Whilst ‘opportunity for rest and relaxation’loads on this factor for repeaters (Table17.3), in the case of first-timers (Table 17.2),it loads on the tertiary attractions, whicharguably represent the least popular attrac-tions. These loadings would seem to supportprevious empirical research on repeat visita-



tion which identified that first-timers are farmore active than repeaters (Oppermann,1996), whilst repeaters are more likely to beseeking relaxation than first-timers (Gitelsonand Crompton, 1984).

Factor 3 mainly represents the tertiaryattractions for which Orlando is less well-known. Despite their quality and abundance,they are overshadowed by the primary andsecondary attractions. Orlando is now tryingto broaden its appeal by emphasizing theseless famous resources, in particular to repeatvisitors (Brodie, 2000). Given that Orlando’snightlife is not a major pull factor for UKholidaymakers in general due to the varietyof, and toll taken by, day-time activities, and

the fact that first-timers are less likely to beseeking relaxation than repeaters, the load-ing of ‘nightlife’ and ‘opportunity for restand relaxation’ on this factor for first-timers(Table 17.2) is not unexpected. The fact that‘bus/trolley service’ loads on this factor forrepeaters only (Table 17.3), although admit-tedly the common variance is low, may beinfluenced by the fact that a statistically sig-nificant (P <0.05) higher number ofrepeaters than first-timers used a car to getaround Orlando.

Factor 4 represents the core attractions.The focus groups highlighted that much ofOrlando’s appeal and fame lay not only inits theme parks but also in its ability to meet


Table 17.2. Factors derived from ‘first-timer’ performance ratings on Orlando’s attributes.


Factor 1: FacilitatorsCustomer service 0.792 0.704Accommodation 0.788 0.689Cleanliness 0.735 0.663Pool 0.717 0.619Safety 0.686 0.551Friendliness of locals 0.570 0.471

Factor 2: Secondary attractionsGoods at bargain prices 0.807 0.717Shopping facilities 0.779 0.664Restaurant VFM 0.760 0.697Variety of restaurants 0.659 0.506

Factor 3: Tertiary attractionsCultural and historic attractions and trails 0.754 0.611

Natural and wildlife attractions and trails 0.743 0.673

Sports facilities 0.606 0.519Nightlife 0.587 0.514Opportunity for rest and relaxation 0.420 0.550

Factor 4: Primary attractionsMany things to see and do 0.848 0.777Something for everyone 0.695 0.692Theme parks 0.621 0.479

Factor 5: Transport plusCar-hire service 0.718 0.610Weather 0.684 0.515Bus/trolley service 0.457 0.374



the diverse needs of large and diversetourist parties, including extended families,over the duration of a typical 2-week holi-day. Given that 90% of the sample was stay-ing in Orlando for 2 weeks and that 70%was travelling in parties of four or more,with no significant differences between first-timers and repeaters, the loading of thesekey destination strengths on a single factoris hardly surprising. Interestingly, the focusgroups identified that part of the appeal oftheme parks for repeaters lay not only inemotional attachment and showing the des-tination to first-timers in the same travellingparty, as per Gitelson and Crompton(1984), but also in theme park augmenta-

tion (Brodie, 2000) since the last visit,which emphasizes the relevance of this issueto both mature destinations as well as thosein earlier stages of the life-cycle.

As in the case of Factor 1, Factor 5 – trans-port plus – would generally seem to representattributes that facilitate the enjoyment ofOrlando’s attractions. Apart from this, thereis little similarity in terms of the loadingsrelated to first-timer and repeater attributeratings; the only common element is ‘car-hireservice’ and an apparent leaning towardstransport. The majority (70%) of respondentsused a hire-car to get around Orlando duringtheir holiday, which is understandable for anumber of reasons: the spread of attractions


Table 17.3. Factors derived from ‘repeater’ performance ratings on Orlando’s attributes.


Factor 1: FacilitatorsAccommodation 0.821 0.706Cleanliness 0.765 0.664Pool 0.761 0.598Safety 0.719 0.613Customer service 0.660 0.632

Factor 2: Secondary attractionsGoods at bargain prices 0.796 0.657Shopping facilities 0.762 0.644Restaurant VFM 0.739 0.644Variety of restaurants 0.664 0.601Opportunity for rest and relaxation 0.477 0.457

Factor 3: Tertiary attractionsNatural and wildlife attractions and trails 0.794 0.674

Cultural and historic attractions and trails 0.777 0.645

Sports facilities 0.608 0.519Bus/trolley service 0.502 0.362

Factor 4: Primary attractionsMany things to see and do 0.771 0.712Something for everyone 0.727 0.649Theme parks 0.669 0.538

Factor 5: Transport plusCar-hire service 0.760 0.618Nightlife 0.514 0.487Road signs that are easy to follow 0.481 0.415Friendliness of locals 0.466 0.505



and activities around the destination; theindependence offered by car travel; the easeof car-hire and price of fuel at the destina-tion. Orlando is a destination where a car isarguably just as necessary to get around toenjoy its attractions, as to get away from them;this may well explain why a statistically signifi-cant (P <0.05) higher number of repeatersthan first-timers used a car to get around,notwithstanding the high level of familiaritywith Orlando. In the case of repeaters (Table17.3), the loading of ‘road signs that are easyto follow’, ‘nightlife’ and ‘friendliness oflocals’ may reflect a greater propensity inrepeaters to explore a destination further(Gitelson and Crompton, 1984).

Regression of tourists’ overall satisfaction onthe factors

The results of the regression of the visitors’overall satisfaction with Orlando against thefactors are given in Tables 17.4 and 17.5, i.e.first-timers and repeaters respectively. Theregression models achieved satisfactory lev-

els of goodness-of-fit in predicting overallsatisfaction as indicated by the multiple cor-relation coefficient (R), coefficient of deter-mination (R2) and F ratio. Firstly, the R valueof independent variables on the dependentvariable is 0.575 (first-timers) and 0.523(repeaters), which shows that the tourists hadhigh satisfaction levels with the factors.Secondly, the R2 values of 0.331 and 0.274suggest that 33% and 27% of the variation inoverall first-timers’ and repeaters’ respectivesatisfaction is explained by the factors. Finally,the F ratio values of 19.169 and 21.911 are sig-nificant at 0.001 indicating that the beta coef-ficients can be used to explain each of thefactors’ relative contribution to the variancein tourist’s overall satisfaction.

In the case of first-timers (Table 17.4),the facilitators carry the heaviest weight intheir overall satisfaction with Orlando; a 1-unit increase in the performance would leadto a 0.382 unit increase in overall satisfac-tion, all other variables being held constant.This may well be due to the high quality ofOrlando on these functional attributes, par-ticularly in comparison with other destina-


Table 17.4. Results of regression of overall satisfaction against first-timer performance ratings onOrlando’s attributes.

Dependent variable: First-time tourists’ degree of overall satisfaction with Orlando(used as a surrogate indicator)

Independent variables: Four orthogonal factors representing the components ofOrlando’s performance

Goodness of fit: Multiple R = 0.575R2 = 0.331Adjusted R2 = 0.314SE = 0.53283

Analysis of variance D.f. Sum of squares Mean square

Regression 4 21.769 5.442 Residual 155 44.006 0.284

F = 19.169Significant F = 0.000

Variable in the equation B SE B Beta T Sig. T

Independent variableFacilitators (Factor 1) 0.246 0.042 0.382 5.820 0.000Secondary attractions (Factor 2) 0.192 0.042 0.298 4.539 0.000Primary attractions (Factor 4) 0.156 0.042 0.243 3.700 0.000Tertiary attractions (Factor 3) 0.123 0.042 0.192 2.917 0.000Constant 4.338 0.041 104.883 0.004


tions, which was highlighted during the pre-liminary primary research. Moreover, givenOrlando’s reputation as a ‘busy’ holiday des-tination amongst UK holidaymakers, forexample in comparison to a more traditional‘3S’ location, these attributes may have amore significant role to play. Secondaryattractions are the second most influentialfactor affecting first-timers’ overall satisfac-tion; a 1-unit increase in the performance ofthese would lead to a 0.298 unit increase inoverall satisfaction.

Since Orlando is renowned for the num-ber and variety of its attractions, especially itsman-made theme parks, it might be expectedthat the primary attractions would make thegreatest contribution to overall destinationsatisfaction. However, the results of theregression identify that these core attractionscarried only the third heaviest weight for first-timers in their overall destination satisfaction.A 1-unit increase in their performance wouldlead to a 0.243-unit increase in tourists’ over-all level of satisfaction, all other variables

being held constant. Given the higher contri-butions of facilitators and secondary attrac-tions, this lower influence may be due tofirst-time visitors’ relative unfamiliarity withOrlando’s high-quality offering in terms ofaccommodation and shopping in comparisonto a greater ‘familiarity’ with its theme parks,particularly due to the latter’s heavy promo-tion. The regression analysis results showedthat the least influential factor on first-timers’overall satisfaction was Orlando’s tertiaryattractions; a 1-unit increase in their perfor-mance would lead to a 0.192-unit increase intourists’ overall level of satisfaction, all othervariables being held constant. Interestingly,‘opportunity for rest and relaxation’ loads onthis factor for first-timers only, which wouldseem to support previous empirical researchon repeat visitation which identified that first-timers are far more active than repeaters(Oppermann, 1997).

For repeaters (Table 5), the secondaryattractions carry the heaviest weight in theiroverall satisfaction with Orlando; a 1-unit


Table 17.5. Results of regression of overall satisfaction against repeater performance ratings onOrlando’s attributes.

Dependent variable: Repeating tourists’ degree of overall satisfaction with Orlando(used as a surrogate indicator)

Independent variables: Five orthogonal factors representing the components ofOrlando’s performance

Goodness of fit: Multiple R = 0.523R2 = 0.274Adjusted R2 = 0.261SE = 0.52338

Analysis of variance D.f. Sum of squares Mean square

Regression 5 30.010 6.002 Residual 291 79.714 0.274F = 21.911Significant F = 0.000

Variable in the equation B SE B Beta T Sig. T

Independent variableSecondary attractions (Factor 2) 0.212 0.03 0.348 6.964 0.000Primary attractions (Factor 4) 0.147 0.03 0.241 4.824 0.000Facilitators (Factor 1) 0.143 0.03 0.234 4.687 0.000Tertiary attractions (Factor 3) 0.0942 0.03 0.155 3.096 0.002Transport (Factor 5) 0.07591 0.03 0.125 2.495 0.011Constant 4.458 0.03 147.773 0.013


increase in the performance of these attrac-tions would lead to a 0.348 increase in overallsatisfaction, all other variables being heldconstant. This is reflected by the fact that theperformance of all attributes within this fac-tor were rated significantly higher (P <0.05)by repeaters than first-timers. Given the‘experiential’ nature of holidays, it is interest-ing that the performance of ‘tangible’ pur-chases and the locations in which they arepurchased make such a contribution. Thismay reflect an attempt to make the experi-ence more tangible and/or be due to theperception that Orlando offers good valuefor money, and even bargains, in terms ofboth food and shopping, which was againidentified in the preliminary primaryresearch. Furthermore, repeaters may havealready identified the best bargains andplaces to shop and eat on previous visits. Theloading of ‘opportunity for rest and relax-ation’ on this factor is interesting for a num-ber of reasons. Given that Orlando is a highlyactive holiday destination for UK holiday-makers, due to the scale and scope of itsattractions and possibly its distance from thebeach, it may be that shopping and diningrepresent crucial opportunities for visitors tore-charge their batteries. Furthermore, theloading may be influenced by the fact thatrepeaters are generally more likely to beseeking relaxation than first-timers (Gitelsonand Crompton, 1984).

Unlike first-timers, Orlando’s primaryattractions carry the second highest weight inrepeaters’ overall satisfaction. Despite theirhigher ranking contribution – a 1-unitincrease in their performance would result ina 0.241-unit increase in overall satisfaction(all other variables being held constant) –their contribution is comparable to that infirst-timers’ overall satisfaction (0.243), whichsuggests some consistency in their role inboth first-time and repeat visitation. This mayreflect the fact that many repeaters return tothe theme parks to show the destination toother first-time group members (Gitelson andCrompton, 1984) and/or because of themepark augmentation.

Facilitators carry the third heaviest weightfor repeaters in their overall satisfaction withOrlando; a 1-unit increase in the perfor-

mance of these attractions would lead to a0.234-unit increase in overall satisfaction, allother variables being held constant. Thiscontrasts with the weighting of this factor forfirst-timers, and may reflect the fact that pre-vious experience has conditioned repeatersto the high standard of these attributes.Tertiary attractions and transport plus makethe lowest contribution to repeaters’ overallsatisfaction. In the case of tertiary attractions,this would seem to be a cause of some con-cern at destination management level, giventhat Orlando is trying to enhance its appealby emphasizing these less famous resourcesto repeat visitors (Brodie, 2000). The contri-bution, admittedly small, of transport plus tooverall satisfaction for repeaters but not first-timers, i.e. a 1-unit increase in the perfor-mance of this factor would lead to a0.125-unit increase in overall repeater satis-faction, all other variables being held con-stant, would seem to support the previousproposal that repeaters make more of aneffort to familiarize themselves with the desti-nation as a whole, and not just its main-stream offerings.

Summary

First-timer and repeater performance ratingson 22 Orlando attributes resulted in five fac-tors: primary, secondary and tertiary attrac-tions, facilitators and transport plus, with astatistically significant difference between thesegments on the secondary attractions – thesingle most influential factor affectingtourists’ overall satisfaction with the destina-tion. The overall satisfaction of first-timersand repeaters was explained by different‘hierarchies’ of factors. First-time visitor satis-faction was explained by a four-factor model,with facilitators and secondary and primaryattractions contributing most to their overallsatisfaction. By comparison, a five-factormodel comprising these same four factorsand an additional transport plus factorhelped explain the overall satisfaction ofrepeat visitors to Orlando; secondary attrac-tions, primary attractions and facilitators car-ried the heaviest weights in repeaters’ overallsatisfaction.



Given that destinations are increasinglybeing challenged to compete for tourists, theyneed to continually build on their strengthsand supplement their offerings in order toboth maintain their appeal and keep the cus-tomer satisfied. In effect, these two key objec-tives for destinations ‘book-end’ the tourist’sholiday decision-making and experience byappealing to tourists in the first instance andsubsequently ‘sending them home happy’, andhopefully ready to return and recommend.

Despite its core reputation as the ‘theme parkcapital of the world’, the regression high-lighted the key role of both facilitators, such asaccommodation and customer service, andsecondary attractions, such as shopping anddining, in visitors’ overall satisfaction withOrlando. Consequently, it would seem thatOrlando is succeeding in keeping its UK mar-ket, comprising first-time and repeat visitorssatisfied, both in general and specifically interms of their main holiday activities.


References

Baker, D.A. and Crompton, J.L. (2000) Quality, satisfaction and behavioural intentions. Annals of TourismResearch 27, 785–804.

Brodie, C. (2000) Telephone interview with UK/European Marketing Director, Visit Florida UK(London), March 1.

Carman, J. (1990) Consumer perceptions of service quality: an assessment of the SERVQUAL dimensions.Journal of Retailing 66, 32–48.

Churchill, G.A. and Surprenant, C. (1982) An investigation into the determinants of customer satisfaction.Journal of Marketing Research 19, 491–504.

Crompton, J.L. and Love, L.L. (1995) The predictive value of alternative approaches to evaluating qualityof a festival. Journal of Travel Research 34, 11–24.


Danaher, P.J. and Mattsson, J. (1994) Customer satisfaction during the service delivery process. EuropeanJournal of Marketing 28, 5–16.

Dorfman, P.W. (1979) Measurement and meaning of recreation satisfaction: a case study in camping.Environment and Behaviour 11, 483–510.

Ekinci, Y., Riley, M. and Chen, J.S. (2000) A review of comparison standards used in service quality and cus-tomer satisfaction studies: some emerging issues for hospitality and tourism research. In: Mazanec,J.A., Crouch, G.I., Brent Ritchie, J.R. and Woodside, A.G. (eds) Consumer Psychology of Tourism,Hospitality and Leisure, Vol. 2. CAB International, Wallingford, pp. 321–332.

Fick, G.R. and Brent Ritchie, J.R. (1991) Measuring service quality in the travel and tourism industry.Journal of Travel Research 29, 2–9.

Gitelson, R.J. and Crompton, J.L. (1984) Insights into the repeat vacation phenomenon. Annals of TourismResearch 11, 199–217.

Gyte, D.M. and Phelps, A. (1989) Patterns of destination repeat business: British tourists in Mallorca,Spain. Journal of Travel Research 28, 24–28.

Jenkins, O.H. (1999) Understanding and measuring tourist destination images. International Journal ofTourism Research 1, 1–15.

Kaiser, H. (1974) An index of factorial simplicity. Psychometrika 39, 31–36.Kotler, P., Bowen, J. and Makens, J. (1999) Marketing for Hospitality and Tourism, 2nd edn. Prentice-Hall,

Englewood Cliffs, New Jersey.Mazursky, D. (1989) Past experience and future tourism decisions. Annals of Tourism Research 11,

pp. 333–344. Meyer, A. and Westerbarkey, P. (1996) Measuring and managing hotel guest satisfaction. In: Olsen, D.M.,

Teare, R. and Gummesson, E. (eds) Service Quality in Hospitality Organisations. Cassell, New York,pp. 185–204.

Oh, H. (1999) Service quality, customer satisfaction, and customer value: an holistic perspective. HospitalityManagement 18, 67–82.

Oliver, R.L. (1980) A cognitive model of the antecedents and consequences of satisfaction decisions.Journal of Marketing Research 17, 460–469.


Oliver, R.L. (1997) Satisfaction: a Behavioural Perspective on the Consumer. McGraw-Hill, London. Oppermann, M. (1996) Visitation of tourism attractions and tourist expenditure patterns – repeat versus

first-time visitors. Asia Pacific Journal of Tourism Research 1, 61–69. Oppermann, M. (1997) First-time and repeat visitors to New Zealand. Tourism Management 18, 177–181.Orlando CVB Research Department (2001) 2000 Overseas Visitors Profile. Orange County Convention and

Visitors Bureau Inc., Florida.Schofield, P. and Fallon, P. (2000) Measuring the importance of critical factors in restaurant service quality

performance evaluation: a triadic perspective. Proceedings of Consumer Satisfaction Research in Tourismand Hospitality Conference. Oxford Brookes University, pp. 159–182.

Stevens, J.P. (1992) Applied Multivariate Statistics for the Social Sciences, 2nd edn. Erlbaum, Hillsdale, NewJersey.

Tabachnick, B.G. and Fidell, L.S. (1996) Using Multivariate Statistics, 3rd edn. Harper and Collins, NewYork.

Tribe, J. and Snaith, T. (1998) From SERVQUAL to HOLSAT: holiday satisfaction in Varadero, Cuba.Tourism Management 19, 25–34.

Watson, A.E., Roggenbuck, J.W. and Williams, D.R. (1991) The influence of past experience on wildernesschoice. Journal of Leisure Research 23, 21–36.

Weber, K. (1997) The assessment of tourist satisfaction using the expectancy disconfirmation theory: astudy of the German travel market in Australia. Pacific Tourism Review 1, 35–45.

Westbrook, R.A. and Oliver, R.L. (1991) The dimensionality of consumption emotion patterns and con-sumer satisfaction. Journal of Consumer Research 18, 84–91.

Yuksel, A. and Rimmington, M. (1998) Customer satisfaction measurement. Cornell Hotel and RestaurantAdministration Quarterly December, 60�70.



Chapter eighteenAristotelian Ethical Values within a Tourism/HospitalityIndustry Context

Glenn F. RossSchool of Business (Tourism Programme), James Cook University, Cairns Campus, Australia2003

Abstract

Ethical beliefs are generally held to be among the more potent moderators of human behaviour. Thoseethical precepts brought to the workplace by potential employees may have a profound influence, notonly upon the outcome of many staff–visitor interactions, but also upon the overall level of successenjoyed by the organization, yet relatively little research has been undertaken in regard to the varietyand intensity of ethical precepts as held by potential tourism/hospitality industry employees. Whilst anincreasing number of people in the tourism/hospitality industry are now aware of the importance of aclear and articulated ethical framework within which to conduct their operations, relatively littleknowledge is currently available with respect to the types of and commitment levels to fundamentalethical precepts that aspirant employees are likely to bring to the industry. Furthermore, it is suggestedthat such ethical beliefs are highly likely to manifest themselves within the context of the serviceencounter, and therefore within the evaluative processes of visitors. This study has sought to examineideal ethical beliefs together with employment context preferences and also perceived visitor, staff andmanagement ethical expectations among a sample of secondary college students in a major Australiantourist destination, many of whom would likely later seek post-secondary college education or immedi-ate employment within the tourism/hospitality industry. It has been found that ethical ideals generallyexceeded those perceived to be held by current tourism/hospitality industry staff and also tourismindustry management; respondent ideals were perceived to be similar to those of visitors, except inregard to precepts such as helpfulness and frankness. Those less likely to perceive a divergencebetween ideal and actual staff friendliness were the ones more likely to favour tourism/hospitality/retail and also tourism/transport employment contexts. Finally, those graduands morelikely to perceive themselves as holding ethical beliefs significantly different from visitors in regard tofrankness were the ones more likely to deem tourism/hospitality industry employment contexts asundesirable. Implications of these various findings are addressed.



Introduction

Boyatzis et al. (2002) have asserted that, incareer decision making, issues such as ethicsare of major importance in individuals’ lives,no matter what the age or the life-stage atwhich that person may be. Argyle (1989) hassuggested that sufficient evidence now existsto conclude the personal values have a stronginfluence over career choice for those seek-ing to enter the workplace. Emsler (1995)points out that career-choice/vocationaldecision making does not belong solely toadult life; young people, including those inthe final years of college, often have devel-oped distinct beliefs and intentions regard-ing the world of work. Moreover, they arelikely to have begun the process of formingintentions as to preferred work contexts, aswell as sets of beliefs in respect of those peo-ple already operating within those contextssuch as employers, employees and customers.Personal values, Emsler further suggests, areimportant components of this career deci-sion-making process, and are likely to accom-pany the individual through the variouspreparatory stages of life and on into theworkplace.

Arnold (2001) advocates the utility of asocio-cognitive psychological perspective inoffering useful insights into such processes,and cites the work of Moss and Frieze (1993),who have found that students’ preferencesamongst job offers could be explained bothby the extent to which offers matched stu-dents’ desired job attributes, and also by theextent to which students’ stereotypes of peo-ple already working within the jobs offeredmatched their self-concept. It is also pointedout by Arnold that career decision makingmight profitably be understood within thecontext of a person–environment interaction;within such a notion, the centrality of the per-sonal values and ethical precepts of the indi-vidual ought not be underestimated.Furnham (1997) also suggests that, within thedomain of vocational choice, person–environ-ment fit models are important: motivation toa particular workplace might best be compre-hended by the fit between the characteristicsof the individual (including their value sys-tems) and their perceptions regarding the

demands of the job and the characteristics ofthe employers, the employees and the cus-tomers. Furnham further asserts that anunderstanding of how individuals conceptual-ize and rate the elements of workplaces isessential in successful management decisionmaking, having long-term effects upon bothproductivity and work satisfaction. Here also,the concept of person–environment fit is ofmajor importance.

Human values

An area of psychology that touches directlyupon issues concerning ethics and morality isthat of human values. A great deal ofresearch has now been accomplished in thisarea, principally within a social psychologicalframework. Perhaps the best-known theoristis this field is Milton Rokeach; within his the-ory values were regarded as preferences fordesirable life states (such as freedom orequality) or as ethical behaviour (such ashonesty or altruism). For each person, valuesare said to be organized within an overall sys-tem, which Rokeach (1973) described as anenduring hierarchy of component valuesalong some continuum of relative impor-tance. Rokeach, moreover, would suggestthat an individual’s value system is generallystable, and may be measured in various ways.Rokeach had people rank order lists ofinstrumental and terminal values accordingto how important such values were regardedas guiding principles within their lives. Thestability within individual value systems isunderstood by many commentators againstthe backdrop of the importance of values tothe integrity of the self; the general stabilityof value systems is held necessary so as toexpress the coherence of the self over timeand across various life situations (Rokeach,1973; Feather, 1975, 1996).

Ethical values, within such a framework,are conceived of as beliefs about desirable orundesirable ways of behaving, or about thedesirability or otherwise of general goals.Such values are more abstract than are atti-tudes, in the sense that they are assumed totranscend specific objects, events and situa-tions. They furthermore are commonly

216 G.F. Ross


regarded as serving an evaluative function:individuals employ these various ethical pre-cepts when they judge outcomes, events andbehaviour, or when they make decisionsupon particular courses of action that typi-cally involve both life goals and those peopleabout them. It is commonly held that theethical values individuals hold are consider-ably fewer in number than are the many spe-cific attitudes that people constantly reveal inrespect to the events, people and environ-ment about them. Commentators such asRokeach (1973), Ball-Rokeach et al. (1984)and Schwartz (1996) would suggest that allethical values may not be regarded as beingequal of importance; such ethical values aresaid to exhibit themselves in each of us at var-ious levels of estimation. Moreover, they arenot, as Feather (1996) and Seligman et al.(1996) have pointed out, cold cognitions,but are typically linked to the affective sys-tem; individuals may feel contented, happy,justified or vindicated when their ethical val-ues find expression or are perceived to befulfilled. They may also, by way of contrast,experience frustration, anger, guilt or confu-sion if such ethical values are frustrated, mis-understood, infracted or rejected.

Ethics and tourism

Within the tourism, hospitality and leisuredomains, there is now an emerging aware-ness of the importance of ethics. Variousstudies have now appeared, relating ethicalunderstandings to specific industry contexts:leisure studies – McNamee et al. (2001);tourism – Fennell (1998); codes of ethics –Coughlan (2000); eco-tourism – Malloy andFennell (1998). Tourism and hospitalityethics education has also seen a surge ofinterest over the last decade or so. BothMartin (1998) and Vallen and Casado (2000)have pointed out that there is a need to pre-pare students to logically and ethically solveindustry dilemmas that they will undoubtedlyface, and that educators need to equip futuretourism and hospitality leaders with the skills,the confidence and the self-esteem to makethe best ethical decisions possible. Therehave been a number of studies examining

ethical conceptualizations and decision mak-ing styles within educational contexts, suchas those reported by Enghagen and Hott(1992), Hall and Enghagen (1991), Kent etal. (1993) and Wheeler (1994). A recentstudy by Stevens (2001) examined theresponses of human resource managers andhospitality students to a variety of ethical sce-narios, involving issues such as theft, racialprejudice, keeping gifts and false accusa-tions. The human resource managers andstudents rated the act of theft the mostunethical, followed by sexual harassment,and then an attempt to obtain proprietaryinformation. Students were revealed to haverated the scenarios as less unethical than didthe human resource managers. Stevens con-cludes by making the point that the studentsdid not view the situations with the samedegree of ethical caution as did the humanresource managers, and suggests that muchis yet to be understood in regard to theprocesses by which students went about mak-ing their ethical decisions.

The derivation of ethical beliefs

Until the latter half of the 20th century,ethics was dominated in large measure by twomajor theories: utilitarianism and Kantian ordeontological philosophy. Deontological the-ory was basically energized by the works ofthe 18th-century philosopher ImmanuelKant; utilitarianism owes its inspiration to thephilosophers Jeremy Bentham and J.S. Mill.In more recent decades, the term virtueethics has come to represent an approach toethics, which highlights ethical character, andis in stark contrast to the approaches thatemphasize rules and duties (deontological)or approaches, which emphasize the conse-quences of a citizen’s actions (utilitarianism).The exponents of virtue ethics would seek toemphasize the character of the person in theunderstanding of any ethical action.Hursthouse (1997) would suggest that virtueethics is both an old and a new theory: old inso far as it reaches back as far as Socrates,Plato and Aristotle, and new in that it hasbeen revivified and remodelled in the secondpart of the 20th century.

Aristotelian Ethical Values in a Tourism Context 217


Virtues, Aristotle suggested, are traits ofcharacter manifested in habitual actions.However, because such an explanation doesnot distinguish between virtues and vices,Pincoffs (1986) has argued that virtues canbe more precisely defined as traits of char-acter manifested in habitual actions that aregood for a person to have. Aristotle sug-gested that many virtues may be identified,and did so within his writings. Rachels(1995, 1998) has distinguished four virtuesfrom Aristotle’s work that are held to be ofprime importance in daily life: courage,generosity, honesty and loyalty to family,friends and close associates. Finally, Rachelsmakes the point the major virtues are man-dated not by social convention, but ratherby basic facts about the common humancondition.

Crisp and Slote (1997) have also heldthat such virtues are traits of character thatare good for people to have; moreover, asRachels has pointed out, the four abovemen-tioned virtues are deemed important for all.Whilst many of the Aristotelian virtues aresaid to fluctuate in intensity and are some-times absent in some members of the popu-lation, the four major virtues cited arecommonly held as important for all people.This research has sought to operationalizeand explore the abovementioned four majorvirtues of courage, generosity, honesty andloyalty among potential employees as theymay mediate tourism industry employmentcontext preferences, and are perceived to beesteemed by visitors, by tourism industrystaff and by tourism management.

Method

Subjects

Four hundred and ninety-three studentsenrolled in years 11 and 12 in a number ofstate high schools from the Cairns region ofnorthern Australia were sampled. Studentswere surveyed during August and September,when many were considering post highschool study or employment options. Ross(1995, 1997, 1998) has found that there isgenerally a high level of interest among sec-ondary school graduates in tourism and hos-pitality industry management employment,with many students being prepared toundergo university/college-level training inorder to achieve these vocational goals.

Measures

Respondents were asked to rate each of thefollowing ethical precepts, according to howimportant they believed them to be rated bytourism industry staff, by tourism industrymanagement and by visitors, and, finally, howimportant they personally regarded them tobe. Each of the four ethical value preceptshave been taken from the notions suggestedby Aristotle, and have also appeared in thewritings of both Rokeach and Feather; fur-thermore, each ethical precept was adaptedto the tourism/hospitality industry workplacecontext, particularly as each precept may beinterpreted within a service quality frame-work (Noe, 1999):

218 G.F. Ross

Important 5 4 3 2 1 Unimportant

Being friendly Being honest Being frank Being helpful

Highly preferred 5 4 3 2 1 Not preferred at all

Mining Manufacturing ForestryHi-technology Rural Light industriesTourism Heavy industries Service industriesTransport Government EducationCommerce/finance Retail Hospitality

Respondents were also requested to rate each of the following work context preferences thus:

Age and gender were also recorded.


Procedure

The survey was distributed among the majorstate high schools in the Far NorthQueensland region of Australia, and adminis-tered to students during class hours by acareers counsellor in each school. The non-response rate was less than 11%.

Results

A principal components factor analysis per-formed on the set of work context prefer-ence ratings revealed three factors witheigenvalues greater than unity. The first fac-tor accounted for 24% of the variance, andloaded on transport and tourism. The sec-ond factor accounted for approximately16% of the variance, and loaded ontourism, hospitality and retail contexts. Thethird factor accounted for over 9% of thevariance, and loaded on the region’s tradi-tional industries such as commerce and gov-ernment. The transport/tourism and the

tourism, hospitality and retail sets ofresponses have been summed and used asfactors in subsequent analyses. It is worthnoting that, because of the dominant posi-tion that the tourism/hospitality industryoccupies within the region surveyed, mosttransportation involves visitors; a great manyof the respondents would, in all likelihood,have had such an understanding in mind.For this reason, the transport factor may beregarded as a type of tourism employmentcontext factor.

A major interest in this study concernedthe divergences (i.e. deviations) as betweenethical ideals and perceived ethical idealsamong tourism/hospitality staff, manage-ment and visitors. Table 18.1 contains means,standard deviations, standard errors andranges associated with divergences betweenthe ideal and perceived staff, managementand visitor ethical expectations. Table 18.2contains a summary of the directionality ofeach of these divergences. These tables revealthe most prominent divergence as betweenrespondents’ own ideals and perceived staff


Table 18.1. Descriptive statistics associated with ethical precept divergences.

Divergence Mean diff. Std dev. Std error Count

Friendlinessideal–staff 2.203 1.511 0.073 429Honestyideal–staff 0.285 1.199 0.057 442Franknessideal–staff 0.557 1.200 0.057 442Helpfulnessideal–staff �0.462 1.484 0.071 442Friendlinessideal–mgmt 0.235 1.146 0.054 446Honestyideal–mgmt 1.077 1.398 0.067 442Franknessideal–mgmt 1.856 1.482 0.071 439Helpfulnessideal–mgmt �0.147 1.585 0.075 442Friendlinessideal–visitors 0.025 1.032 0.049 447Honestyideal–visitors �0.011 1.185 0.057 440Franknessideal–visitors 0.419 1.140 0.054 442Helpfulnessideal–visitors �0.381 1.451 0.069 441


friendliness ideals. The helpfulness ethicalprecept, however, evidenced a differentresponse: the perceived staff ideal was ratedhigher than the respondent’s ethical ideal. Inregard to ideal-management divergences,honesty and frankness emerged as beingthought of as more important than did man-agement. For ideal–visitor ethical diver-gences, frankness was deemed moreimportant for the respondents than they wereperceived so to be by the visitors; helpfulness,though, was found to be reverse: visitors wereperceived to rate this more highly than wouldthe respondents.

Paired t-test analyses have been applied topairs of ethical ratings, and also to pairs ofethical divergences; Table 18.3 containspaired t-tests for ethical precepts, whereasTable 18.4 contains paired t-tests for ethicaldivergence comparisons. From Table 18.3 itcan be seen that most pairs of ratings evi-denced significant differences; those that didnot involved the ideal–visitor analysis forfriendliness, the ideal–visitor analysis for hon-esty, the staff–visitor analysis for franknessand the staff–visitor analysis for helpfulness.Table 18.4 reveals that the only non-signifi-cant analysis involved the comparisonideal–staff vs. ideal–visitor for the helpfulnessethical precept; for helpfulness, therefore,the differences between the respondent–staffdivergence and the respondent–visitors diver-gence did not emerge as significant.

Multiple regression analyses have beenemployed in this study so as to examine therelative predictive power of each divergence,using the two tourism-related work contextfactors as criterion variables. Two significantfunctions were so formed, both involvingideal–staff divergences. Table 18.5 reveals thefriendliness ideal–staff divergence to signifi-cantly predict tourism/hospitality/retailemployment preference, with the directional-ity of the standardized coefficient suggestingthat those respondents less likely to perceivean ideal–staff divergence as being more likelyto favour this type of employment. Table 18.6reveals a similar result, revealing those stu-dents who saw their own friendliness idealsand those of tourism industry staff membersas similar being the ones more likely to favoura tourism/transport work context.

Finally, Kruskal–Wallis analysis of variance(ANOVA; by ranks) procedures have beenemployed so as to further explore the frank-ness ethical precept in relation to tourismindustry employment preference. Table 18.7reveals that lower preferences for tourism/transport context employment to be signifi-cantly associated with higher estimations of thevisitor frankness ideal, whereas lower levels ofthe tourism/transport employment interestwere associated with mid-range and higher lev-els of the visitor frankness precept. For thefrankness ideal–visitor ethical precept diver-gences, however, lower preferences for the

220 G.F. Ross

Table 18.2. Directionality of ethical precept divergences.

Ethical precept Mean difference Directionality of divergence

Ideal–perceived tourism staff idealFriendliness 2.203 Ethical ideal clearly exceeds staff idealHonesty 0.285 Ethical ideal exceeds staff idealFrankness 0.557 Ethical ideal exceeds staff idealHelpfulness �0.462 Perceived staff ideal exceeds own ideal

Ideal–perceived tourism management idealFriendliness 0.235 Ethical ideal exceeds mgmt idealHonesty 1.077 Ethical ideal clearly exceeds mgmt idealFrankness 1.856 Ethical ideal clearly exceeds mgmt idealHelpfulness �0.147 Perceived mgmt ideal exceeds own ideal

Ideal–perceived visitor idealFriendliness 0.025 Little differenceHonesty �0.011 Little differenceFrankness 0.419 Ethical ideal exceeds visitor idealHelpfulness �0.381 Perceived visitor ideal exceeds own ideal



Table 18.3. Paired t-tests performed on ethical precept ratings.

Divergence Mean diff. D.f. t-value P-value

Friendliness ethical divergencesIdeal–staff 2.203 428 30.192 <0.0001Ideal–mgmt 0.235 445 4.339 <0.0001Ideal–visitor 0.025 446 0.504 0.6143Staff–mgmt �1.969 445 �28.287 <0.0001Staff–visitor �2.201 437 �31.334 <0.0001Mgmt–visitor �0.204 454 �3.799 0.0002

Honesty ethical divergencesIdeal–staff 0.285 441 4.998 <0.0001Ideal–mgmt 1.077 440 16.185 <0.0001Ideal–visitor �0.011 439 �0.201 0.8407Staff–mgmt 0.781 473 11.972 <0.0001Staff–visitor �0.286 454 �5.271 <0.0001Mgmt–visitor �1.088 451 �17.200 <0.0001

Frankness ethical divergencesIdeal–staff 0.557 441 9.751 <0.0001Ideal–mgmt 1.856 438 26.247 <0.0001Ideal–visitor 0.419 441 7.716 <0.0001Staff–mgmt 1.229 474 16.554 <0.0001Staff–visitor �0.115 452 �1.812 0.0707Mgmt–visitor �1.396 449 �18.781 <0.0001

Helpfulness ethical divergencesIdeal–staff �0.462 441 �6.536 <0.0001Ideal–mgmt �0.147 441 �2.851 0.0407Ideal–visitor �0.381 440 �5.515 <0.0001Staff–mgmt 0.297 463 4.731 <0.0001Staff–visitor 0.091 449 1.443 0.1497Mgmt–visitor �0.241 451 �3.720 0.0002

Table 18.4. Paired t-tests performed on ethical divergence comparisons.

Divergence Meancomparisons differences D.f. t-value P-value

Friendliness divergence comparisonsI–S vs. I–M 1.981 425 28.228 <0.0001I–S vs. I–V 2.188 424 31.086 <0.0001I–M vs. I–V 0.197 441 3.653 0.0003

Honesty divergence comparisonsI–S vs. I–M �0.781 437 �11.522 <0.0001I–S vs. I–V 0.305 435 5.477 <0.0001I–M vs. I–V 1.090 434 17.112 <0.0001

Frankness divergence comparisonsI–S vs. I–M �1.297 436 �17.113 <0.0001I–S vs. I–V 0.139 438 2.175 0.0301I–M vs. I–V 1.422 435 19.000 <0.0001

Helpfulness divergence comparisonsI–S vs. I–M �0.303 438 �4.736 <0.0001I–S vs. I–V �0.068 437 �1.073 0.2839I–M vs. –V �0.239 439 3.641 0.0003

I–S, ideal–staff; I–M, ideal–management; I–V, ideal–visitor.


transport context were associated with ahigher divergence as between respondents’ideals and perceived visitor ideals; higher pref-erences for this work context were associatedwith lower divergences. Thus Table 18.8

reveals that those students, who were morelikely to perceive themselves as holding differ-ing values in regard to frankness to those ofvisitors, were the ones more likely to view thiswork context as undesirable.

222 G.F. Ross

Table 18.5. Multiple regression analyses of ideal–tourism staff ethicaldivergences, using tourism/hospitality/retail employment factor as thecriterion variable.

Summary statistics table

D.f. Sum of sqs Mean sqs f-value

Regression 4 300.687 75.172 7.940, P <0.0001Residual 402 3806.030 9.468Total 406 4106.717

Standardized coefficient table

Coefficient Std error Std coeff. t-value

Intercept 10.906 0.289 10.906 37.678, P < 0.0001Friendlinessideal–staff �0.535 0.101 �0.255 �5.300, P < 0.0001Honestyideal–staff 0.084 0.133 0.032 0.633, P < 0.5273Franknessideal–staff �0.241 0.133 �0.090 �1.816, P < 0.2032Helpfulnessideal–staff 0.134 0.105 –0.063 1.274, P < 0.2032

Table 18.6. Multiple regression analyses of ideal–tourism staff ethicaldivergences, using the transport factor as criterion variable.

Summary statistics table

D.f. Sum of sqs Mean sqs F-value

Regression 4 71.990 17.997 5.256, P = 0.0004Residual 401 1373.114 3.424Total 405 1445.103

Standardized coefficient table

Coefficient Std error Std coeff. t-value

Intercept 6.358 0.174 6.358 36.496, P <0.0001Friendlinessideal–staff �0.278 0.061 �0.223 �4.573, P = <0.0001Honestyideal–staff 0.002 0.080 0.002 0.031, P = 0.9755Franknessideal–staff �0.009 0.080 �0.006 �0.112, P = 0.9106Helpfulnessideal–staff 0.041 0.063 0.032 0.649, P = 0.5169


Discussion

This study has revealed a number of interest-ing and useful findings in regard to the rela-tionship between ethical precepts, ethicalperceptions and tourism industry employ-ment preferences. First of all, it has foundthat graduands appeared to conceptualizetourism industry employment in two separatedomains: that of tourism transport employ-ment, and that of tourism, hospitality and

retail employment. Thus employment in air-lines, in bus lines, in tour operations, in rail,in coach lines, in taxi companies, in limousineservices and such like, may be identified as adifferent domain to that of other tourism,hospitality and retail employment. Whilstthese two separate employment domains werenot found to be different in regard to ethicalpredictors, it is worth noting that, to thisgroup of potential employees, the two con-texts are identifiably distinct.


Table 18.7. Kruskal–Wallis ANOVA (by ranks) of the perceived visitorsfrankness ethical precept, by the transport employment preference factor.

Summary statistics factor

D.f. 2Number of groups 3Number of ties 5H 9.663P value 0.008H corrected for ties 11.452P corrected for ties 0.0033

Mean ranks table

Transportpreference factor Count Sum of ranks Mean ranks

Lower preference 145 33,576.0 231.559Mid range pref. 95 20,945.5 220.479Higher preferences 182 34,731.5 190.832

Table 18.8. Kruskal–Wallis ANOVA (by ranks) of the franknessideal–visitor ethical precept divergence by the transport employmentpreference factor.

Summary statistics factor

D.f. 2Number of groups 3Number of ties 9H 6.164P value 0.0459H corrected for ties 7.427P corrected for ties 0.0244

Mean ranks table

Transportpreference factor Count Sum of ranks Mean ranks

Lower preference 177 38,959.5 220.11Mid range pref. 91 18,513.5 203.445Higher preferences 141 26,372.0 187.035


Ethical precepts and also ethical diver-gences among ethical precept ratings havebeen found to vary considerably in this study;this was so for the four precepts examined andalso for the various ethical precept ratings.Ethical ideals were generally ranked higherthan perceived tourism staff ratings and alsoperceived tourism management ratings; thistendency was particularly marked in regard tothe friendliness ethical divergence, betweenrespondents’ ideal and perceived tourism staffideal. Thus respondents clearly believed thatthey were, or would be, more friendly towardvisitors than current tourism industry person-nel. In contrast, these potential staff regardedthemselves as being similar in terms of friendli-ness to that expected by visitors. In terms ofhelpfulness and frankness, however, they per-ceived themselves as differing distinctly fromvisitors: visitors were seen to expect more help-fulness than the students believed they shouldoffer, whereas the respondents declared anethical obligation to be somewhat more frankthan was deemed to be ideal, at least in theperceived view of the visitors.

The friendliness ethical precept ideal–staffdivergence has been found to be a significantpredictor of tourism industry employmentpreference, for both the tourism transportpreference factor and also the tourism, hospi-tality, retail preference factor. Those individu-als more likely to entertain beliefs whereinethical precepts and tourism industry staffethical precepts in regard to friendliness evi-denced little or no difference were the onesmost likely to elect for either employmentcontext. Thus those people who regardedthemselves as being similar, at least in regardto this major ethical belief, were the ones whomost desired future employment in thetourism industry in its various manifestations,whereas those for whom this friendliness gulfwas wider, did not.

The two ethical precepts that did evoke dif-ferences as between respondents and visitorswere helpfulness and frankness. Visitors wereperceived to require more help than manyindividuals were willing to offer; such findingsdo raise concerns in respect of service. Possiblymany of these school-leavers had developed adistorted view of what it was that visitorsrequire in terms of service. Alternatively, a

number of this group may be unsuited fortourism industry employment, or indeed anyservice industry employment, wherein pleas-ant and satisfying staff–visitor interactions arean essential element of the product. Tourismis, at core, an experience, and negative visitorexperiences in regard to staff service willgreatly detract from that. It would thereforeseem essential that the industry, in its recruit-ing and also in its overall media representa-tions, be mindful of such issues.

Similar sentiments might also beexpressed in respect of the frankness find-ings. The greater the estimation of this frank-ness precept, and also the greater thedivergence as between respondent ideal andperceived visitor expectation, the more grad-uands were found to eschew tourism industryemployment contexts. It is possible thatfrankness is here being interpreted by someas a type of bluntness, an assertiveness thattakes little account of the sensitivities, theconfusion and even the vulnerabilities experi-enced by some visitors. If this is so, then suchindividuals do well to avoid this industry andits clients. Those for whom frankness, as apersonal value and also as a visitor expecta-tion, is much more aligned may be more ser-vice-oriented and understanding of visitors inunfamiliar environments; such respondentsmay realize that visitors do often value a can-dour and openness displayed by staff whenthey temper their frank utterances withinsight and courtesy.

Limitations to this study ought now beaddressed. The study has not primarily con-cerned itself with the behaviour of consumers;it has, however, involved ethical expectationsthat tourists are believed to embrace. Thestudy, moreover, did not involve those individ-uals presently employed within thetourism/hospitality industry. Rather, it wasdirected toward those secondary college grad-uands, many of whom would soon enter intosuch employment or enrol in post-secondaryor university courses, a major emphasis ofwhich would be preparation for such employ-ment. Future research on this topic area,based upon samples from post-secondary/uni-versity and also industry employment contextsmight now afford some insight into how matu-ration, education and employment may medi-

224 G.F. Ross


ate change upon these ethical belief struc-tures. Finally, this study has not included anybehaviourally grounded outcome measures;specific tourism/hospitality job seeking, jobacquisition, job acceptance and job-socializa-tion success variables might now be of somebenefit in understanding the relationshipsbetween ethical values and tourism/hospital-ity industry work context behaviours.

This study has established clear associa-tions in regard to Aristotelian ethical pre-cepts, ethical perceptions and expectations,and tourism industry employment contextpreferences. Findings here build upon the

understandings of ethical precepts as theymediate important decision making withintourism industry arenas; whereas some find-ings from this study give cause for optimismin terms of future service quality levels, otherfindings may not be so encouraging.Members of the tourism industry, particularlyin their recruiting, need a clear focus uponthe nature of ethics and ethical expectationsamong those whom they choose to employ, ifthey are to maintain and enhance their ethi-cal standing as corporate citizens and alsoguarantee service quality to those visitors thatthey seek to serve.


References

Argyle, M. (1989) The Social Psychology of Work. Penguin, London.Arnold, J. (2001) The psychology of careers in organisations. In: Cooper, C. and Robertson, I. (eds)

Organisational Psychology and Development. John Wiley & Sons, Chichester, UK.Ball-Rokeach, S., Rokeach, M. and Grube, J.W. (1984) The Great American Values Test. Free Press, New York.Boyatzis, R., McKee, A. and Goleman, D. (2002) Reawakening your passion for work. Harvard Business

Review, April, 87–94.Crisp, R. and Slote, M. (eds) (1997) Virtue Ethics. Oxford University Press, New York.Coughlan, R. (2001) An analysis of professional codes of ethics in the hospitality industry. Hospitality

Management 20, 147–162.Emsler, N. (1995) Socialisation for work. In: Collett, P. and Furnham, A. (eds) Social Psychology at Work:

Essays in honour of Michael Argyle. Routledge, London.Enghagen, L.K. and Hott, D.D. (1992) Students’ perceptions of ethical issues in the hospitality and

tourism industry. Hospitality Research Journal 15, 41–50.Feather, N.T. (1975) Values in education and society. Free Press, New York.Feather, N.T. (1996) Values, deservingness and attitudes toward high achievers. In: Seligman, C., Olsen,

J.M. and Zanna, M.P. (eds) The Psychology of Values. L.E.A., Mahwah, New Jersey.Fennell, D.A. (1998) Tourism and applied ethics. Tourism Recreation Research 25, 59–69.Furnham, A. (1997) The Psychology of Behaviour at Work. Taylor & Francis, London.Hall, S. and Enghagen, L. (1991) Ethics and hospitality: two tested, full credit models for teaching. Annual

CHRIE Proceedings, 252–253.Hursthouse, R. (1999) On Virtue Ethics. Oxford University Press, Oxford.Kent, W., Lian, K., Kahn, M. and Anene, J. (1993) College’s hospitality programs: perceived quality. Cornell

Hotel and Restaurant Administration Quarterly 34, 90–95.McNamee, M.J., Sheridan, H. and Buswell, J. (2001) The limits of utilitarianism as a professional ethic in

public sector leisure policy and provision. Leisure Studies 20, 173–197.Malloy, D.C. and Fennell, D.A. (1998) Ecotourism and ethics: moral development and organizational cul-

tures. Journal of Travel Research 36, 47–56.Martin, L. (1998) Ethical principles for the hospitality curriculum. Journal of Hospitality and Tourism

Education 10, 22–25.Moss, M.K. and Frieze, H. (1993) Job preferences in the anticipatory socialisation phase: a comparison of

two matching models. Journal of Vocational Behavior 42, 282–297.Noe, F. (1999) Tourist service Satisfaction: Hotel, Transportation and Recreation. Sagamore, Champaign, Illinois.Pincoffs, E.L. (1986) Quandaries and Virtues: Against Reductivism in Ethics. University of Kansas Press,

Lawrence, Kansas.Rachels, J. (1995) The Elements of Moral Philosophy. McGraw-Hill, New York.Rachels, J. (ed.) (1998) Ethical Theory. Oxford University Press, New York.Rokeach, M. (1973) The Nature of Human Values. Free Press, New York.


Ross, G.F. (1995) Tourism/hospitality management employment interest as predicted by job attributes andworkplace evaluations. Tourism Recreation Research 20, 63–71.

Ross, G.F. (1997) Tourism/hospitality industry employment acquisition strategies, higher education prefer-ences and the work ethic. Managing Leisure: An International Journal 2, 82–93.

Ross, G.F. (1998) The Psychology of Tourism. Hospitality Press, Melbourne.Schwartz, S. (1996) Value priorities and behavior: applying a theory of integrated value systems. In:

Seligman, C., Olsen, J.M. and Zanna, M.P. (1996) (eds) The Psychology of Values. L.E.A., Mahwah, NewJersey.

Seligman, C., Olsen, J.M. and Zanna, M.P. (eds) (1996) The Psychology of Values. L.E.A., Mahwah, NewJersey.

Stevens, B. (2001) Hospitality ethics: responses from human resource directors and students to seven ethi-cal scenarios. Journal of Business Ethics 30, 233–242.

Vallen, G. and Casado, M. (2000) Ethical principles for the hospitality curriculum. Cornell Hotel andRestaurant Administration Quarterly 41, 44–51.

Wheeler, M. (1994) The emergence of ethics in tourism and hospitality. Progress in Tourism, Recreation andHospitality Management 6, 46–56.

226 G.F. Ross


Chapter nineteenThe Role of Expressive and Instrumental Factors inMeasuring Visitor Satisfaction

Muzaffer Uysal1 and John Williams2

1Department of Hospitality and Tourism Management, Virginia Polytechnic Institute and StateUniversity, 355 Wallace Hall, Blacksburg, VA 24061-0429, USA; 2Department of Hotel,Restaurant, Institution Management and Dietetics, Kansas State University, 103 Justin Hall,Manhattan, KS 66506-1404, USA

Abstract

This study describes overall satisfaction as a function of instrumental and expressive factors. The objectiveof the study is accomplished by testing whether instrumental and expressive attributes are distinct behav-ioural indicators that could better predict visitor satisfaction. Furthermore, the study tests whether visitortypes based on motivation for travel moderate the relationship between instrumental and expressive attrib-utes. The findings of the study revealed partial support that expressive and instrumental factors collectivelymight be predictors of overall satisfaction or dissatisfaction in general. However, the findings reveal thatvisitor types based on motivation for travel moderate the relative importance of instrumental and expres-sive factors. Empirical studies of this nature may be of help to destination marketers and planners tounderstand the complexity of satisfaction as one of the elements of visitation behaviour. Actual and poten-tial markets can use these types of studies to develop appropriate communication materials that wouldincorporate the relative importance of destination features as perceived.

Introduction

The demand side of tourism harbours thenotion of value that tourists attach to theirleisure and recreation experiences. This valuemay take on either the perceived importanceof destination attributes, or service and bene-fit(s) to be received and the transaction valueof the service being rendered (Heskett et al.,1997). Managers of public and private out-

door recreational areas need confirmationfrom the visiting public that the facilities, ser-vices and programmes generally provided aresatisfactory (Noe, 1999; Schofield, 2000). Theobjective of this study is to sharpen scalingtechniques by testing whether the perceivedimportance of instrumental and expressiveattributes are distinct behavioural indicatorsthat could better predict visitor satisfaction.Expressive indicators involve core experi-



ences representing the major intent of an act(exhibits, interpretation of programmesoffered, experiences with recreation activities,etc.), whereas instrumental indicators serveto act as facilitators toward achieving thatdesired end (parking, rental services,restrooms, etc.; Noe 1987). Thus, the mainelements of satisfaction – expressive andinstrumental attributes – in tourism settingscan be examined within the context of atourism system representing two major com-ponents of the market place, namely demandand supply.

Although these distinctions are usuallyused to specifically define characteristics of arecreational situation that may affect satisfac-tion, the visitor, through reasons for travel orunfulfilled expectations, may also signifi-cantly alter judgements of satisfaction aboutthat situation.

The development of a working model isbased upon an outgrowth of earlier researchby Noe (1987). This model assumes a directapproach for determining satisfaction andmakes a theoretical distinction betweeninstrumental and expressive indicators of satis-faction. In a subsequent study, Noe and Uysal(1997) also examined the theoretical distinc-tion between instrumental and expressiveindicators of satisfaction with respect to differ-ent outdoor settings. One of the major con-clusions of their study was that the relativeimportance of both instrumental and expres-sive indicators may show variation from site tosite. Examination of the differences in thenature of different sites may help to explainthe different relative importance of destina-tion attributes in creating desired leisureexperiences. Most recently, Uysal and Noe(2003) provided comparisons of the results ofsurveys previously gathered from tourists visit-ing National Parks and tried to explore howattributes in a tourist situation are specifiedand perceived. Two issues dominated theirresearch, namely identifying indicators orattributes of satisfaction and specifying a satis-faction model. From the review of the parksurveys, they concluded that both expressiveand instrumental attributes of sites and desti-nations could collectively and independentlycontribute to satisfaction and the practicalusefulness of the findings improves when data

pertain to specific programme components,rather than complex global programmes orgeneral issues. Space is thus responsible forthe unequal (or equal) spatial distribution ofvisitation. Therefore, the place dependency ofsatisfaction should be of great concern inmeeting the expectations of tourists.Situational and structural differentiation ofproducts and services would pose additionalchallenges for providers to be more proactivein facilitating the enjoyment of product offer-ings. Reviews of the case studies from the parksurveys also revealed that the most satisfyingare the expressive character of the site(s) inquestion. However, it is important to remem-ber that instrumental factors can have moreof an effect on perceptions of dissatisfaction.Therefore, it is important to recognize thatthe existence and nature of resources asattractions would still inherently have theirown constraints, regardless of the level ofimportance that visitors may attach to destina-tion attributes. Providers would still strive tomaintain a co-aligned strategy between instru-mental performance and visitor expectations.

In marketing, Swan and Combs (1976)define instrumental performance as themeans to an end or the evaluation of thephysical product, whereas the expressiveattribute was the end in itself or the psycho-logical interpretation of a product. In socialaction theory, both concepts are treated asnecessary for human action. Both are goal-directed with the instrumental being morecognitively oriented, whereas the expressive ismore emotional or feeling oriented (Noe,1999). Swan and Combs (1976) also assertthat satisfaction can be produced onlythrough the expressive activities. The evalua-tive mode of behaviour is also associated withexpressive acts within the context of socialaction theory. On the other hand, Noe(1987) found that expressive indicators of sat-isfaction forming core recreational experi-ences were more salient in explaining generalsatisfaction. Czepiel and Rosenberg (1974)would consider these factors that ‘truly moti-vate and contribute to satisfaction’, whereasthe instrumental are maintenance factors,which, if absent, create dissatisfaction. Fromthis argument, it is clear that facilities andattractions may possess the duality of expres-

228 M. Uysal and J. Williams


sive and instrumental roles that may comple-mentarily produce satisfaction. For example,in analysing the effects of the instrumentaland expressive attributes on satisfaction, aLISREL path model was applied to analysethe variables so as to best select ‘the richestand most parsimonious model… to explainthe satisfaction process’ (Jurowski et al., 1995,p. 56). In testing the efficacy of the variousmodels, the final model resulting from thisresearch implies that instrumental andexpressive satisfiers work together to produceoverall satisfaction.

In short, expressive indicators involve coreexperiences representing the major intent ofan act, in this case seeking a satisfactory out-door experience in a park (sightseeing, camp-ing, hiking a natural trail, floating a river,etc.). Instrumental indicators serve as actionsor behaviours toward facilitating that desiredend (rental services, restrooms, concessionservices, etc.) (Uysal and Noe, 2003). In ourtesting of the expressive–instrumental model,it is still reasonable to speculate that instru-mental factors could also contribute to satis-faction/dissatisfaction, since so few studieshave measured these attributes. It is clearfrom the review of related research that someexpressive attributes are the behaviouralresults of inner emotional state. These attrib-utes are the essence of travel motivation inthe first place, representing the demand sideof the equation. Therefore, the current studyalso posits that visitor types may also moder-ate the relationship that may exist betweenexpressive and instrumental attributes and

satisfaction. In the study, motivation for travelis used to define visitor types.

The current working model describesoverall satisfaction as a function of instrumen-tal and expressive importance of destinationattributes. Furthermore, the model tests themoderation effect visitor types may have onthe relationship. The major question raisedin this research asks to what extent does theabove expressive, instrumental attributes, andthe visitor types relate to overall satisfaction.Figure 19.1 depicts the conceptual model ofthe study. Using a structural equation model(SEM), the study examines the structural,causal relationships among the constructs.Hypothetically, both expressive and instru-mental factors influence satisfaction, which isthen moderated by visitor types based onmotivation. In this model, expressive andinstrumental factors are considered to be theexogenous variables (i.e. those that are notpredicted by any other variables in themodel); satisfaction is the endogenous vari-able (the one that is affected by the others).The relationship between the importance ofexpressive and instrumental factors and satis-faction depends on visitor types.

Research Methods

Study coverage

This study was initially conducted to investi-gate general consensus of potential visitors fora nature-based resort destination. Visitors were

Expressive and Instrumental Factors in Visitor Satisfaction 229

Satisfaction

Visitortype

Expressivefactors

Instrumentalfactors

Fig. 19.1. Conceptual model.


defined in terms of distance from the destina-tion, within the radius of a 3–5 h drive, andthose who have taken at least one pleasure tripaway from home for at least 2 nights. Thisstudy included places from the states ofVirginia, West Virginia, North Carolina,Maryland and Washington, DC, area. Theanalysis of the study focused on the mostrecent pleasure trip of the potential visitors.This portion of the study was subcontracted toNational Family Opinion (NFO). NFO main-tains a panel of over 50,000 households for itsTravels America. The telephone surveys, 15–20min each, were conducted over a period of 1month in summer 2002. A total sample of 409was generated. The study used a proportion-ally stratified random sampling techniquebased on the distribution of visitors by origin.

Variables/questionnaire

The questionnaire consisted of four sections:(i) general travel behaviour; (ii) most recentpleasure trip; (iii) lodging options for devel-opment and construction; and (iv) respon-dent characteristics. In the first section,respondents were asked to provide their gen-eral travel behaviour. This section of the sur-vey included importance of travel attributes(motivations), importance of travel value,importance of destination attributes, prefer-ences of pleasure trip. Since perceived desti-nation attribute is used as a key predictor inexplaining part of the variance in satisfaction,a five-point Likert scale for rating the per-ceived importance of destination attributesfrom ‘not at all important’ to ‘very important’was developed. The second section of thequestionnaire was specifically designed toassess market potential for development andconstruction of different lodging options(cabin on its own, single stand-alone unitwith limited meals and activities, condo-minium/apartment type and time-sharearrangement for a limited number of days in ayear) at the selected resort destination.Section three focused on the travel behaviourof the potential visitors in relation to theirmost recent pleasure trip. This section con-tained questions on number of pleasure trip,primary destination, activity participation,

travel group, travel decision time, moneyspent, information sources used, familiaritywith competing destinations and intention tovisit, and travel worth and satisfaction. Travelworth was measured on a five-point Likert-type scale from 1 ‘not quite worth’ to 5 ‘defi-nitely well worth’. Satisfaction was alsomeasured on a five-point Likert-type scale:1 ‘not satisfied at all’ to 5 ‘very satisfied’.

Analysis

The study used SEM empirically to test therelationships between the constructs. Thisincludes the classification of the constructsinto dependent (endogenous) and indepen-dent (exogenous) constructs. The propertiesof the items of the constructs (one endoge-nous, three exogenous) in the proposedmodel and the general hypothesis were testedusing the LISREL 8 structural equation analysispackage (Joreskog and Sorbom, 1996) withthe maximum likelihood (ML) method ofestimation in conjunction with the two-stageprocess. The relationship between observableindicators and the theoretical constructs con-stitutes the measurement part of the model,and the theoretical relationships between theconstructs constitute the structural part ofthe model (Joreskog, 1993). Confirmatoryfactors analysis (CFA) for each construct withcomposite reliability and variance-extractedestimate was also conducted. Then, the unidi-mensionality of each construct and the over-all measurement of four constructs were alsochecked. Fourteen out of 27 ‘observed indica-tors’ were retained and subsequently used fortesting the proposed model.

Identification of visitor types

In order to examine the moderation effect ofvisitor types based on motivation on the rela-tionship between expressive and instrumentalimportance of destination attributes and satis-faction, the study first identified the visitortypes based on motivations for travel andthen used the resultant visitor types in thesubsequent LISREL analysis to measure themoderation effect of the visitors’ type on the



relationship between expressive and instru-mental factors and satisfaction. The identifi-cation of the visitor types followed afactor–cluster segmentation approach. Thisprocedure was carried out in three steps. Instep 1, 12 travel motivation items were factoranalysed to see the underlying dimensions ofthe destination attributes and in step 2, thestudy used cluster analysis to group visitorsbased on the factor scores. Finally, a discrimi-nant analysis was performed to understandthe degree to which the destination attributescould distinguish the responses of visitorsbelonging to each cluster.

The cluster analysis based on factor scoresof the motivation items identified two types ofvisitors. These visitor types are called type Iand type II. Type I included 184 respondents(50.8% of the sampled visitors); type II con-tained 178 respondents (49.2% of the sam-pled visitors). Forty-seven cases wereeliminated from this analysis due to pair-wisedeletion of missing values. Overall, 97.0% ofthe grouped cases were correctly classified.

Type I visitors seem to place more valueand importance on such destination attrib-utes as visiting new places, fun and enjoyment,and less importance and value on visitingfamiliar places, having time myself, and visitingfriends and relatives. These visitors were classi-fied as ‘Novelty and fun & enjoyment seek-ers’. Visitors in type II seem to place morevalue and importance on taking a carefully andcompletely planned trip, how much money theyspend, having a restful relaxing trip, visitingfriends and relatives, and togetherness and close-ness with family or friends. These visitors wereclassified as ‘Familiarity and comfort seekers’.The difference between the two with respectto destination attributes (motivation items)lies in the degree of importance they attachto such attributes.

Results

Testing the proposed model

A total of 409 samples were collected andanalysed. Missing values, outliers and distribu-tion of all measured variables were examinedto purify the data and reduce systematic

errors. Prior to LISREL analyses, exploratoryfactor analysis (EFA) was performed for pur-poses of reducing the number of variables inboth expressive and instrumental constructs.Four instrumental (INTS) and eight expres-sive indicators were derived. The expressivehad two sub-constructs, labelled as Naturalsetting (EXP1) and Activities (EXP2);Satisfaction (SAT) had two indicators. Then,the included items within a factor were calcu-lated to create a composite factor.

Subsequently, these composite factors weretreated as indicators to measure a construct.This procedure may help to decrease multi-collinearity or error variance correlationsamong indicators in the confirmatory factoranalysis of the measurement model. Sucherrors should be avoided as much as possiblein SEM procedures (Bollen, 1989). The prop-erties of the four research constructs (threeexogenous – INTS, EXPR1 and EXP2 – andone endogenous – SAT) were tested with a LIS-REL procedure of SEM (Joreskog and Sorbom,1996). The ML method of estimation and thetwo-stage testing process were adopted.

Measurement model

Confirmatory measurement models should beevaluated and re-specified before measure-ment and SEM models are examined simulta-neously (Anderson and Gerbing, 1988). Thus,before testing the measurement model overall,each construct in the model was analysed sepa-rately. Four indicators of exogenous variablesfor the instrumental, four for each of the twoexpressive constructs and two for visitor satis-faction are determined from the measurementmodel. This measurement model describedthe nature of the relationship between latentconstructs and the manifest indicators thatmeasured those latent constructs. In this study,although the chi-squared test was significant,other conventional goodness-of-fit indices indi-cated that the overall measurement model wasstill acceptable, in that the proposed model fitthe collected data with a sample size of 409 (χ2

= 146.38, d.f. = 71, P = 0.00, GFI = 0.95, RMR =0.047, RMSEA = 0.051, AGFI = 0.93, PNFI =0.64, CFI = 0.97, IFI = 0.95 and CN = 268.47;AGFI, adjusted goodness-of-fit index; CFI,



comparative fit index; CN, critical sample size;GFI, goodness-of-fit index; IFI, incremental fitindex; PNFI, parsimony normed fit index;RMR, root mean square residual; RMSEA, rootmean square error of approximation). Theproposed measurement model fits the samplemarginally well.

After assessing the overall model, the psy-chometric properties of each latent constructwere evaluated separately through examiningthe completely standardized loading, error vari-ance, the construct reliability and the varianceextracted. The construct reliability of all fourconstructs was close, and exceeded the recom-mended level of 0.70 with the exception of thesatisfaction construct. Thus, it can be said thatthe psychometric properties of each respectivelatent construct, especially for the purpose ofthis research, is acceptable (Table 19.1).

Structural equation model

Having assessed the measurement model, thetheoretical model was examined with threegamma paths. The model that was estimated

with three gammas from three latent con-structs, showed a non-significance result of thechi-squared test. The results of goodness-of-fitindices exhibited a similar pattern to those forthe initial theoretical model, as well as indi-cated better fits for all measures (GFI = 95,RMSR = 0.05, AGFI = 0.91, NNFI = 0.96, PNFI =0.61, CFI = 0.97 and IFI = 0.97). Consequently,the review of the squared multiple correlationsof the structural model explained around 10%of the variance in tourist satisfaction. Figure19.2 presents the results of the tested structuralmodel. A closer examination of the structuralpaths of the model reveals that there are twosignificant paths coefficients. The path coeffi-cient from Expressive 1 to satisfaction is 0.10 (P <0.05) and from the instrumental to satisfac-tion is �0.16 (P <0.05). The findings support atleast in this study that the instrumental andpart of the expressive (the setting) influencesatisfaction. However, the direction of the influ-ence is positive for the setting part of theexpressive and ‘negative’ for the instrumental,suggesting that the absence and or poor per-formance of ‘the maintenance factors’ couldinfluence satisfaction negatively.


Table 19.1. Overall CFA for the measurement model.

Completelystandardized Construct and

loading indicator ErrorConstruct and indicators (t-value) reliability variance

Expressive constructsThe setting (EXP1) 0.80 0.49

Beautiful country site (X1) 0.85 (19.72) 0.72 0.28Natural features and scenic wonders (X2) 0.82 (19.95) 0.67 0.25Mountains (X3) 0.69 (12.55) 0.47 0.84Quaint towns and villages (X5) 0.68 (13.50) 0.46 0.69

Activities (EXP2) 0.80 0.62Canoeing (X6) 0.72 (18.14) 0.52 0.25Bicycling (X7) 0.78 (16.37) 0.47 0.46Hiking and backpacking (X8) 0.81 (16.21) 0.66 0.51Snow skiing (X9) 0.46 (10.84) 0.21 0.52

Instrumental construct (INST) 0.84 0.74Convenience to home or being easy to get to (X12) 0.53 (8.23) 0.28 1.10A good value for a vacation trip (X13) 0.59 (11.63) 0.35 0.58Ease of getting around (X14) 0.68 (14.57) 0.46 0.35High-quality service and accommodation (X15) 0.63 (12.07) 0.40 0.60Visitor satisfaction (SAT) 0.43 0.72Worth time and effort (Y1) 0.32 (2.47) 0.10 0.56Overall satisfaction (Y2) 0.45 (7.63) 0.20 0.21


Testing the moderation effect of visitor types

The basic premise of the moderating effect isthat responses to variations in the relationshipbetween expressive and instrumental factorsand satisfaction depend on the visitor type asdefined by motivation items for taking thetrip. In order to test the moderation (interac-tion) effect of the delineated two visitor types,two steps were followed. The first stepinvolved a ‘multiple–group’ solution in whichLISREL derived parameter estimates for eachvisitor type separately without constraintsacross the visitor types as well as a measure ofpooled goodness-of-fit of the model for bothvisitor types I and II considered simultane-ously. In the second step, the path coefficientswere estimated for each visitor type with anacross-group constraint imposed to reflect theinteraction effect (Jaccard and Wan, 1996).Based on the size of the difference in thevalue of chi-squared changes from the basemodel to the non-constrained to the con-strained solutions, a decision about the inter-action effect could be made. Table 19.2provides the statistical comparisons of thevisitor-type models for the interaction effectand Table 19.3 presents the SEM statisticalresults of the estimated path coefficients withtheir associated t-values for visitor types I andII. Both Tables 19.2 and 19.3 support that the

relationship between instrumental factors andpart of the expressive factors (the setting) andsatisfaction is strongest for visitor type II(‘Familiarity and comfort seekers’).

Concluding Comments

The findings partially support that at least inthis study the instrumental and part of theexpressive (the setting) may influence satis-faction. The study explained around 10% ofthe variance in tourist satisfaction. However,the direction of the influence is positive forthe setting part of the expressive and ‘nega-tive’ for the instrumental, suggesting that theabsence and or poor performance of ‘themaintenance factors’ could influence satisfac-tion negatively. The results also reveal enoughevidence to suggest that visitor types maymoderate the relationship between instru-mental factors and part of the expressive fac-tors (the setting) and satisfaction. Themoderation effect is strongest for visitor typeII (‘Familiarity and comfort seekers’). It isclear from the study reported that instrumen-tal attributes also influence satisfaction alongwith at least part of the expressive.

Some expressive attributes are the behav-ioural results of inner emotional state. Theseattributes are the essence of travel motivation


X1

X2

X4

X5

X5

X6

X7

X8

X12

X13

X14

X150.60

0.28

0.25

0.84

0.69

0.25

0.46

0.51

0.52

1.10

0.58

0.35

EXP1

EXP2

INST

SAT

0.10

–0.16

0.03

Y1

Y2

0.56

0.96

Fig. 19.2. SEM results of the hypothesized model.


in the first place, representing the demandside of the equation. The responses todemand side or expressive attributes, includ-ing benefits sought at the destinations, wouldthen naturally represent the supply side oftravel experience. Therefore, the instrumen-tal are maintenance attributes without whichone may not achieve some degree of satisfac-tion on a measurement scale from the experi-ence. The quality and availability of tourismsupply resources are a critical element inmeeting the needs of the ever changing andgrowing tourism market. It is important thatdestinations monitor visitors’ satisfaction withfacilities, programmes and services in orderto maintain a sustained and expanding busi-ness. Empirical studies of this nature may beof help to destination marketers and plannersto understand the complexity of satisfactionas one of the elements of visitation behaviour.Actual and potential markets can use thesetypes of studies to develop appropriate com-munication materials that would incorporatethe relative importance of destination fea-tures as perceived. The results of the concep-tual model partially confirm the existence ofdual satisfaction factors. These findings arealso consistent with the results reported byNoe (1987), Floyd (1993), Jurowski et al.

(1995) and Noe and Uysal (1997). It appearsthe factors representing a judgement aboutfacilities and services may not be as importantto overall satisfaction for the visitor type of‘Novelty and fun & enjoyment seekers’ asthey are for the visitor type of ‘Familiarity andcomfort seekers’. In this current study, theinstrumental factors appear to be moreimportant than expressive elements for thevisitor. The poor performance of site-specificinstrumental factors may affect satisfactionnegatively, while the expressive part of expec-tations from a site may affect satisfaction posi-tively. Thus, the expressive part ofexpectations may also contribute to overallsatisfaction. These results imply that visitorsatisfaction at sites designed for visitor experi-ences would not necessarily depend uponexpressive elements such as the setting itself,hiking and interpretive programmes, whilevisitor satisfaction in areas designed for sight-seeing, and providing activities may be some-what more dependent on the provision ofinstrumental elements such as restrooms,ease of accessibility, high-quality services andaccommodation. This observation suggeststhat further research is also needed to exam-ine the effect of development stages of desti-nation sites on satisfaction.


Table 19.2. Moderation effect of visitor type based on motivation.

Change in Chi-squared D.f. Change in d.f. chi-squared Prob.

Base model 225.76 142 – – –Expressive (EXP1) 228.13 143 1 2.37 0.123Expressive (EXP2) 225.82 143 1 0.06 1.00Instrumental (INTS) 228.52 143 1 2.76 0.096

Table 19.3. SEM statistical results of visitor types I and II.

Visitor type Coefficient t-value

Type I (Novelty and fun & enjoyment seekers)Expressive (EXP1) – Satisfaction (SAT) �0.11 1.13Expressive (EXP2) – Satisfaction (SAT) 0.05 0.74Instrumental (INTS) – Satisfaction (SAT) �0.05 �0.43

Type II (Familiarity and comfort seekers)Expressive (EXP1) – Satisfaction (SAT) 0.16 2.05a

Expressive (EXP2) – Satisfaction (SAT) 0.03 0.41Instrumental (INTS) – Satisfaction (SAT) �0.26 �2.55a

aSignificant at 0.05 and better probability level.


Tourists form expectations of a destina-tion based upon advertising and promotionalcampaigns, past experience and word-of-mouth that, in turn, may influence demandfor tourism destinations. The quality of theservice, and the quality of the facility and itsperceived value, also directly affect the qual-ity of recreation and tourism experiencesand thus the level of future demand.Furthermore, the level of satisfaction that thetourist feels is also dependent upon the abil-

ity of the destination to deliver the type ofexperience, which it has marketed as a func-tion of its facilities and programmes (Ryan,1995). The degree to which managers con-trol the commonality between instrumentalperformance and the psychological interpre-tation of such performance will eventuallyinfluence satisfaction and the correspondingdevelopment of successful programmes, facil-ities, and management and monitoringmechanisms.


References

Anderson, J.C. and Gerbing, D.W. (1988) Structural equation modeling in practice: a review and recom-mended two-step approach. Psychological Bulletin 103, 411–423.

Bollen, K. (1989) Structural Equations with Latent Variables. John Wiley & Sons, New York.Czepiel, J.A. and Rosenberg, L.J. (1974) The study of consumer satisfaction. AMA Educators’ Proceedings.

American Marketing Association, Chicago, pp. 119–123.Floyd, M. (1993) Recreation Research Survey of Cape Lookout National Seashore. Final report submitted to the

national Park Service, Texas A&M University, College Station, Texas. Heskett, J.L., Sasser, W.E. and Schlesinger, L.A. (1997) The Service Profit Chain. The Free Press, New York.Jaccard, J. and Wan, C.I. (1996) LISREL Approaches to Interaction Effects in Multiple Regression. Sage,

Thousand Oaks, California.Joreskog, K.G. (1993) Testing structural equation models. In: Bollen, K.A. and Long, J.S. (eds) Testing

Structural Equation Models. Sage Publications, Newbury Park, California, pp. 294–316. Joreskog, K.G. and Sorbom, D. (1996) LISREL VIII Manual. Scientific Software, Inc., Mooresville, Indiana.Jurowski, C., Cumbow, M., Uysal, M. and Noe, F. (1995) The effects of instrumental and expressive factors

on overall satisfaction in a park environment. Journal of Environmental Systems 2491, 47–67.Noe, F. (1987) Measurement specification and leisure satisfaction. Leisure Sciences 9, 163–172.Noe, F. (1999) Tourist Service Satisfaction: Hotel, Transportation, and Recreation. Sagamore Publishing,

Champaign, Illinois.Noe, F. and Uysal, M. (1997) Evaluation of outdoor recreational settings: a problem of measuring user sat-

isfaction. Journal of Retailing and Consumer Services 4, 223–230. Ryan, C. (1995) Researching Tourist Satisfaction: Issues, Concepts, and Problems. Routledge, London.Schofield, P. (2000) Evaluating Castlefield Urban Heritage Park from the consumer perspective: destina-

tion attribute importance, visitor perception, and satisfaction. Tourism Analysis 5, 183–189. Swan, J. and Combs, L. (1976) Product performance and consumer satisfaction. Journal of Marketing

Research 40, 25–33.Uysal, M. and Noe, F. (2003) Satisfaction in outdoor recreation and tourism settings. In: Laws, E. (ed.) Case

Studies in Tourism Marketing. Continuum Publisher, London, pp. 140–158.


Chapter twentyProfiling Airline Web Users: a Segmentation Approach

Joseph S. Chen1 and Seyou Jang2

1Department of Hospitality Management, International University of Applied Sciences,Mülheimer Strasse 38, D-53604 Bad Honnef, Germany; 2School of Tourism, Sejong University,98 Gunja-dong, Gwangjin-ku Seoul 143-747, South Korea

Abstract

Due to the paucity of literature dealing with Internet usages, this research attempts to find users’ prefer-ences to airline websites with a segmentation approach. The measurement of the users’ preferences isaccomplished through a two-stage, analytical scheme embracing both qualitative and quantitative meth-ods. Consequently, 16 website-related attributes are developed for the mail survey that yields 328 usefulsamples. Two distinct segments named Bargain Seeker and Utilitarian are collectively determined by a clusteranalysis, the importance rankings and mean scores. Bargain Seeker tends to be younger female with a col-lege degree; has an annual income lower than $60,000; and uses the Internet for over 2 h daily. In the finalanalysis, marketing implications along with suggestions for future studies are rendered.

Introduction

In light of the recent revolution on informa-tion technology, tourism establishments areteeming to capitalize on the new inventionsto make their business operations more prof-itable. With this approach, firms are able tocapture and tailor current customers’ needsrapidly; businesses could attract new cus-tomers effectively and, therefore, makethemselves more competitive – especially inan increasingly challenging market environ-ment. Among the innovations, the Internet isthe most powerful application that has grad-ually changed the landscape of today’s busi-

ness operations, such as inventory control,human resources management and salesforecasting. Undoubtedly, the Internet willgreatly influence consumers’ life routines inthe years to come. With the integration ofthe Internet, businesses could obtain oppor-tunities to dominate the future marketplace.

In response to the technology revolution,airlines have fast implemented new strategiesto add the Internet into their business mod-els. It is evidenced that the new informationmedia renders airlines vast advantages toaugment their market share, so as to reducethe operational costs. For example, it is likelythat air passengers may want to access the



Internet during long-distance flights. Airlinescould provide Internet services on someselected flights to increase the demand.

Purposes of the study

In the last few years, a growing number ofconsumer studies has paid great attention toonline operations. Although the contribu-tions to the body of literature on e-commercehave increased, empirical investigations per-taining to airline operations are still lacking.To fill the void of past literature concerningInternet user behaviours, this researchattempts first to measure Internet users’ pref-erences to airline websites, given the fact thatunderstanding customers’ needs and wantswill promise business a long-term success.Furthermore, the current research aims todetermine various market segments accord-ing to preferences for airline websites.

Specifically, four research questions aredeveloped, as follows: (i) can Internet usersbe grouped into various mutually exclusivesegments by preference attributes? (ii) whatare the major attributes discriminating thesegments? (iii) what are the distinct demo-graphic traits and consumption patterns ineach segment? (iv) do the segments differ sig-nificantly among different levels of demo-graphic characteristics (e.g. income) andInternet usage (e.g. hours of use)?

Market segmentation studies

A plethora of segmentation investigationshas been conducted in the last threedecades. The aim of these studies is to detectunique sub-groups from a heterogeneouspopulation that could help marketing man-agers develop cost-effective solutions to drivethe market demand. In review of recenttourism studies, a battery of works has beenfound that utilized a board range of attrib-utes, as bases or descriptors, to identify theunderlying segments.

These segmentation bases and descriptorshave largely accentuated the characteristics ofsocio-demographics (Hudson, 2000; Lee andLee, 2000; Muller and Cleaver, 2000), psycho-

graphics (McCleary and Choi, 1999;Yannopoulos and Rotenberg, 1999; Kim et al.,2000; May et al., 2001), decision-behavioural(Chen and Gursoy, 2000; Mok and Iverson,2000), geographic (Dolnicar et al., 1999),activity-specific (Choi et al., 1999; Koo et al.,1999; Nicholson and Pearce, 2000), product-specific (Chen, 2000; Shoemaker, 2000) andpersonal interests (Olsen et al., 2000).Irrespective of the selection of segmentationattributes, it is importance to recognize thatthe resultant segments should be practical,meaningful and actionable (Kolter, 1988) forthe development of marketing strategies.

Methodology

This study adopted a two-stage approach toaccomplish the data collection. In the firststage, the researchers invited a group of indi-viduals to report their use preferences andpast experiences concerning travel-relatedwebsites. Instead of focusing on the opinionsof those who had accessed airline Web pagesbefore, the researchers posited to understandthe preferences of both current and potentialusers. Therefore, individuals browsing travel-related Web information were included in thefirst stage of study. Along with the importantWeb preferences found in the extant litera-ture, this study incorporated the results fromthe focus group survey into a close-endedquestionnaire that was used in the secondstage of measurement.

In the second stage, a mail survey was con-ducted. Regarding the sampling frame, sincethe acquisition of a customer database fromairlines was rather difficult, a mailing list wasacquired from a commercial mailing broker.The study questionnaire included a prefer-ence scale and the measurements of socio-demographic characteristics. The preferencemeasurement embraced 16 attributes basedon a five-point Likert-type scale. These attrib-utes were: (i) secured customers information;(ii) promotional offerings; (iii) critical travelinformation; (iv) simplicity of Web design; (v)booking confirmation; (vi) multi-languageservices; (vii) guest feedback; (viii) flexibleschedule; (ix) speedy transaction; (x) aircraftinformation; (xi) customer support services;

238 J.S. Chen and S. Jang


(xii) price comparisons among differentoptions; (xiii) guest history; (xiv) linkage toother service providers; (xv) special mealrequest; and (xvi) connecting airport infor-mation.

Before the deployment of the mail survey,tourism professors and graduate students firstscreened the draft questionnaire. A pre-testof the questionnaire with a sample of 35Internet users was then carried out toenhance the validity and reliability of thepreference measurement. Consequently, 1500questionnaires were mailed to travellers whohad used the Internet as a tool to retrieve air-line-related information. From the mail sur-vey, this study collected 328 usablequestionnaires.

In the data analysis procedure, to identifymutually exclusive groups, a non-hierarchicalcluster technique – k-means cluster analysis –on 16 preference attributes was used.Meanwhile, all outliers were removed byusing Box plots, which identified outlier casesof which values were replaced by the meanvalue of the particular attributes diagnosed bythe plot, before the employment of the clus-ter analysis.

To understand the unique traits of eachcluster, two supplementary tests – logisticregression and the chi-squared test – wereconducted. A stepwise logistic regression wasperformed to find the differences in prefer-ence attributes between/among the clusters;a series of chi-squared tests was used to deter-mine if there were any differences in demo-graphic characteristics between/among theclusters. If any demographic characteristicdiffered significantly between/among theclusters, the logit analysis was then used tofurther derive predictive parameters on thesignificant variables.

Results and Discussions

Resultant clusters

From the results of the non-hierarchical clus-ter analysis, a two-cluster solution appeared tobe appropriate in explaining the heterogene-ity of the study population. Therefore, theanalysis further partitioned the study popula-

tion into two segments. Table 20.1 shows theresults of the descriptive analysis of 16 prefer-ence attributes. Both clusters I and II onlyhave one attribute (secured customers infor-mation) with an average rating over ‘Agree (=4)’. The finding suggests that all Web usersare seriously concerned about the Internetsecurity issues. It is not surprising that theInternet users viewed websites as an unsafecommunication channel to supply personalinformation, since online purchasing is arather new business activity.

In Table 20.1, cluster I, ‘promotional offer-ings (ranking = 2)’, ‘critical travel informa-tion (raking = 3)’, and ‘simplicity of Webdesign (ranking = 4)’ are also top-rated pref-erence attributes. Information regarding pro-motional products and services seem to beimportant figures that the respondents fromcluster I are distinctively longing for.Interestingly, the above three attributes arenot the highly rated preferences by cluster II.It appears that the respondents in cluster IIare apt to look for ‘speedy transaction (rank-ing = 2)’, ‘customer support services (ranking= 3)’, and ‘price comparisons among differ-ent options (ranking = 4)’.

About 60% of respondents from cluster Iare females, which account for only 40% ofthe respondents of cluster II. Regarding mar-tial status, both the majority of cluster I (=54%) and II (= 63%) are married. Forty-oneper cent of the respondents from cluster Ihave an annual household income between$30,001 and $45,000; however, in cluster II,the individuals with a similar income levelrepresent 35% of the cluster. Most respon-dents (= 48%) in cluster I use the Internet for2–3 h daily; however, the majority (= 57%) ofthe respondents in Cluster II surf theInternet for 1–2 h only.

Differences in Web preferences

Beyond the examination of preference rank-ings, this study also tested the significant dif-ferences in use preferences between theclusters by using a stepwise logistic regressionanalysis. As Tables 20.2 and 20.3 indicate, thefinal logistic model’s classification rate ishigh (93%), using 11 preference attributes

Profiling Airline Web Users 239



Table 20.1. Descriptive analysis on airline Web use preferences.

Ranking (Mean)

Web use preferences Cluster 1 Cluster 2

Secured customers information 1 (4.17) 1 (4.67)Promotional offerings 2 (3.83) 12 (2.90)Critical travel information 3 (3.62) 11 (3.28)Simplicity of Web design 4 (3.28) 10 (3.35)Booking confirmation 5 (3.27) 6 (3.44)Multi-language services 6 (3.21) 15 (2.48)Guest feedback 7 (3.19) 5 (3.52)Flexible schedule 8 (3.18) 8 (3.38)Speedy transaction 8 (3.18) 2 (3.80)Aircraft information 10 (2.98) 16 (2.34)Customer support services 11 (2.83) 3 (3.76)Price comparisons between different options 12 (2.79) 4 (3.66)Guest history 13 (2.75) 7 (3.43)Linkage to other service providers 14 (2.68) 14 (2.79)Special meal request 15 (2.39) 9 (3.37)Connecting airport information 16 (2.24) 13 (2.85)

All attributes are measured by a five-point scale (1= least important; 2 = less important; 3 = moderate;4 = very important; 5 = most important).

Table 20.2. Classification results from logistic regression.

Predicted group membership

Actual cluster No. of cases 1 2

Cluster 1 156 141 (90.38%) 15 (9.62%)Cluster 2 172 8 (4.65%) 164 (95.35%)Percentage of cases classified correctly: 93%

Table 20.3. Logistic regression of airline Web preference attributes.

Preference attributes Coefficient estimate Wald statistic P

Promotional offerings �4.57 36.94 0.000Aircraft information �4.01 37.34 0.000Flexible schedule 3.76 29.72 0.000Multi-language services �3.49 25.88 0.000Meal request 2.66 23.13 0.000Guest history 2.59 25.08 0.000Connecting airport info 2.17 27.00 0.000Price comparison �1.63 11.91 0.000Guest feedback �1.62 13.44 0.000Simplicity of Web design 1.57 5.83 0.016Speed transaction 1.51 14.38 0.000χ2 = 6.404, d.f. =11, P = 0.0011

All attributes are measured by a five-point scale (1= least important; 2 = lessimportant; 3 = moderate; 4 = very important; 5 = most important).


including: (i) promotional offerings; (ii) air-craft information; (iii) flexible schedule; (iv)multi-language services; (v) meal request;(vi) guest history; (vii) connecting airportinformation; (viii) price comparisons; (ix)guest feedback; (x) simplicity of Web design;and (xi) speedy transaction as the base forthe analysis.

The largest differences between the clustersoccur in the category of promotional offer-ings. The result indicates that cluster I is moresensitive to promotional offerings (r = �4.57),and cluster II is inclined to emphasize on flexi-ble schedule (r = 3.76) and guest history (r =2.66). When consolidating the results from thepreference rankings, the study finds that therespondents in cluster I are looking for bar-gain deals from the websites, whereas theInternet users from cluster II desire ‘quality’websites that could help them make reserva-tions in a user-friendly fashion.

In the review of the above variationsregarding use preferences and importancerankings, this study labels cluster I as Bargain

Seeker and cluster II as Utilitarian. The respon-dents of the two clusters also reveal undistin-guished preferences comprising: (i) linkageto other service providers; (ii) confidentialityof personal information; (iii) customer sup-port features; (iv) booking confirmation; and(v) regional information.

Differences in demographic traits and Webusages

Table 20.4 depicts the results of chi-squaredtests, which were applied to the clusters inorder to establish demographic differencesand use variations. The tests established sig-nificant differences in ‘gender’, ‘age’, ‘educa-tion’, ‘income’, and ‘Web use hours’ betweenthe resultant segments. Moreover, four post-hoc analyses (logit modelling) were furtherconducted in the bid to determine whichindividual levels of demographic andInternet-usage variables differed significantlybetween the clusters.


Table 20.4. The profiles of two resultant clusters.

Bargain Seeker Utilitarian

Variables (%) (%) Chi-squared P

Gender 13.37 0.000Male 49.4 69.2Female 50.6 30.8

Age 21.71 0.00025 and Less 16.7 9.926–35 69.2 54.136–45 5.1 16.2Over 45 9.0 19.8

Education 12.98 0.005High school or below 7.1 7.0Some college 26.3 23.8College 55.7 43.0Graduate 10.9 26.2

Income 22.76 0.000Below $30,000 25.0 14.0$30,000–45,000 41.0 34.9$45,001–60,000 28.2 27.9Above 60,000 5.8 23.3

Web use hours 24.25 0.000Less than 1 h 14.1 14.51–1.99 h 32.1 57.02–2.99 h 48.1 25.63 h and over 5.8 2.9


In Table 20.5, the results of logit analysesdepict the categorical differences among age,education, income and Web use hours. As forage, the results exhibit that the age groups of36–45 and over 45 have negative values. Thisindicates that respondents over 35 are likelyto be Utilitarian. Conversely, the youngerrespondents are inclined to be Bargain Seekers.It suggests that younger users look for bar-gain airfares when using the Internet as analternative way of acquiring additional airlineinformation and making a reservation.

Among the categories of ‘education’, theattributes of ‘some college’, ‘college’, and‘graduate’ differ significantly between theclusters. Bargain Seeker appears to have a lowereducation level. The results lead to a sugges-tion that Utilitarian is more likely to have agraduate degree. Regarding the income lev-els, all income categories vary between theclusters. Bargain Seeker tends to earn less than$60,000 while Utilitarian receives over $60,000annually. Lastly, the logit model on ‘Web usehours’ shows that four levels of use hours dif-

fer significantly. Bargain Seeker has an inclina-tion to utilize the Web for over 2 h on a dailybasis, whereas Utilitarian is unlikely to use theInternet for less than 2 h.

Conclusion

Based on 16 preference attributes, two distinctclusters of respondents are derived. A logisticregression analysis further manifests that theclusters possess a different degree of needregarding 11 service preferences. According tothe dissimilarities between the derived clusters,the two mutually exclusive groups of Internetusers are labelled Bargain Seeker and Utilitarian,respectively. Notably, the respondents of bothsegments share similar preferences to fiveWeb-related features. With the revelations, air-lines could further utilize these preferenceattributes to improve their service strategies.For example, the five indistinctive preferenceattributes shall be considered critical servicefigures attracting all groups of Web users.


Table 20.5. Logit models of age, education, income and Web use hours.

Parameter Estimate Chi-squared P

1. Model on ageIntercept �0.29 22.57 0.0025 and Less 0.81 56.19 0.0026–35 0.54 53.27 0.0036–45 �0.86 45.09 0.00Over 45 �0.48 20.08 0.00

2. Model on educationIntercept �0.12 4.79 0.00High school or below 0.14 1.13 0.28Some college 0.22 6.74 0.05College 0.38 27.74 0.00Graduate �0.75 57.25 0.00

3. Model on incomeIntercept �0.15 8.93 0.00Below $30,000 0.73 64.99 0.00$30,000–45,000 0.31 18.68 0.01$45,001–60,000 0.16 4.49 0.03Above $60,000 �1.22 110.84 0.00

4. Model on Web use hoursIntercept 0.17 6.66 0.00Less than 1 h �0.21 3.87 0.041–1.99 h �0.75 78.46 0.002–2.99 h 0.45 26.26 0.003 h and over 0.51 8.64 0.00


It is enlightening to find two distinct seg-ments according to use preferences. BargainSeeker might represent those mainly utilizingthe Web as an information channel to findlow-fare air tickets. This segment consists ofyounger, less-educated individuals who have alower annual income and, however, are will-ing to spend more time on the Web. A mar-keting implication of this segment is thatairlines might develop mailing lists of the seg-ment that help the airlines effectively informBargain Seeker of their promotional offeringson a regular basis. Retrospectively, forUtilitarian, airlines might consider establish-ing a customized hyperlink for Utilitarian thatwill provide a speedy transaction, better cus-tomer supports and sophisticated price com-parison functions.

From segment comparisons, the studyfinds an interesting implication that individ-uals spending more hours on the Internet,such as Bargain Seeker, are sensitive to promo-tional offerings in general. The above phe-nomena might be attributed to the frequentexposure to online offerings from all e-com-

merce sectors. For future e-commerce stud-ies, it might be desirable to measure therelationships between the length of expo-sure to promotional offerings and the inten-tion to purchase.

This study confirms that Internet securityis the major concern of airline Web users. Itimplies that the reputation of airline websitesmay largely depend on the service providers’ability to safeguard customer informationsaved in the Web servers. From a practicalpoint of review, developing a secured data-base system would gain the competitive edgefor online sales. Moreover, since informationsecurity is one of the major issues frequentlyworrying both service providers and con-sumers across all e-commerce sectors, furtherresearch on the impacts of Internet securityon consumers’ purchase intention might alsobe worth conducting. Lastly, it is important torecognize a distinct limitation regarding thesize of the study sample. Due to the short-coming, this study is unable to cross-validatethe resultant segments. A further study onsuch an issue seems vital.


References

Chen, J.S. (2000) Norwegians’ preferences to U.S. lodging facilities: a market segmentation approach.Journal of Travel and Tourism Marketing 9, 68–82.

Chen, J.S. and Gursoy, D. (2000) Cross-cultural comparison of the information sources used by first-timeand repeat travelers and its marketing implications. International Journal of Hospitality Management 19,119–203.

Choi, W., Tsang, K., Wai, M.C. and Tsang, C.K.L. (1999) Activity based segmentation on pleasure travelmarket of Hong Kong private housing residents. Journal of Travel and Tourism Marketing 8, 75–97.

Dolnicar, S., Grabler, K. and Mazanec, J.A. (1999) Analyzing destination images: a perceptual chartingapproach. Journal of Travel and Tourism Marketing 8, 43–45.

Hudson, S. (2000) The segmentation of potential tourists: constraint differences between men andwomen. Journal of Travel Research, 38, 363–368.

Kim, S., Crompton, J.L., Botha, C. and Kim, S.S. (2000) Responding to competition: a strategy forSun/Lost city, South Africa. Tourism Management 21, 33–41.

Koo, L.C., Tao, F.K.C. and Yeung, J.H.C. (1999) Preferential segmentation of restaurant attributes throughconjoint analysis. International Journal of Contemporary Hospitality Management 11, 242–250.

Kotler, P. (1988) Marketing Management: an Analysis, Planning, Implementation, and Control. Prentice Hall,Englewood Cliffs, New Jersey.

Lee, C. and Lee, C.K. (2000) A comparative study of Caucasian and Asian visitors to a cultural expo in anAsian setting. Tourism Management 21, 169–176.

May, J.A., Bastian, C.T., Taylor, D.T. and Whipple, G.D. (2001) Market segmentation of Wyoming snow-mobilers. Journal of Travel Research 39, 292.

McCleary, K. and Choi, B.M. (1999) Personal values as a base for segmenting international markets.Tourism Analysis 4, 1–17.

Mok, C. and Iverson, T.J. (2000) Expenditure-based segmentation: Taiwanese tourists to Guam. TourismManagement 21, 299–305.


Muller, T.E. and Cleaver, M. (2000) Targeting the CANZUS baby boomer explorer and adventurer seg-ments. Journal of Vacation Marketing 6, 154–169.

Nicholson, R. and Pearce, D.G. (2000) Who goes to events: a comparative analysis of the profile character-istics of visitors to four South Island events in New Zealand. Journal of Vacation Marketing 6, 236–253.

Olsen, W.K., Warde, A. and Martens, L. (2000) Social differentiation and the market for eating out in theU.K. International Journal of Hospitality Management 19, 173–190.

Shoemaker, S. (2000) Segmenting the mature market: 10 years later. Journal of Travel Research 39, 11–26.Yannopoulos, P. and Rotenberg, R. (1999) Benefit segmentation of the near-home tourism market: the

case of Upper New York State. Journal of Travel and Tourism Marketing 8, 41–55.



Chapter twenty-oneTowards More Thorough Data-driven Segmentation inTourism: a Tracking Framework for Exploring SegmentDevelopment

Sara DolnicarSchool of Management, Marketing & Employment Relations, University of Wollongong,Wollongong, NSW 2522, Australia

Abstract

Market segmentation has become a standard concept in tourism marketing. A priori and a posteriori(data-driven, post hoc) segmentation approaches enjoy high popularity among both practitioners andresearchers. In order to optimize the market segmentation strategy, it is not only necessary to identify rele-vant market segments, describe them, evaluate the match between corporate or destination strengths andsegment needs, but to understand how segments develop over time. This knowledge is typically accountedfor when a priori segments are used. In the case of a posteriori segments, however, such trend tracking isneglected. In this chapter, a simple tracking framework is presented that allows testing of a posteriori seg-ment developments over time on the basis of identical consecutive guest surveys. It comprises the follow-ing steps: (i) definition of the anchor period; (ii) computation of a data-driven market segmentationsolution; (iii) characterization of market segments; (iv) assignment of data from other periods to theanchor segment solution; (v) testing of distribution changes; (vi) testing of changes in background vari-ables; and (vii) validation of results. The framework is flexible with regard to methods applied at each stepand – through validation of explorative findings by means of repetition – allows insight into market struc-ture from multiple perspectives.

Introduction

Market segmentation has become a standardconcept in strategic marketing. Alongside thewide use of a priori (Mazanec, 2000) segmen-tation approaches, splitting individuals on thebasis of pre-defined criteria, data-driven (or aposteriori, Mazanec 2000) approaches havegained increased popularity in the past

decades among researchers attempting toderive market segments on the basis of surveyinformation using various segmentationbases, e.g. as behavioural or psychometricinformation (Middleton, 1988; Smith, 1995;Myers, 1996; Lilien and Rangaswamy, 1998).While it is common to monitor the develop-ment of a priori segments over time (e.g.development of tourism from certain coun-



tries of origin, different age groups, families,etc.), data-driven approaches are typicallyconducted at one point of time only. Thismight be due to the fact that data-driven seg-mentation per se is an exploratory conceptand studying multiple time periods in such anexploratory manner (when even the oneperiod case is full of possible pitfalls; Punjand Stewart, 1983; Ketchen and Shook, 1996;Baumann, 2000) would further increase com-plexity and opacity and decrease reliability ofresults. The fact, however, that tracking a pos-teriori segments is extremely uncommon isstrongly supported by the literature survey inthe field of business studies in general con-ducted by Baumann (2000) and the subse-quent analysis limited to tourismsegmentation studies (Dolnicar, 2002).Among 47 data-driven segmentation publica-tions within the field of tourism, not a singlestudy reports on investigations over time.

The advantages of being able to trace aposteriori segment trends in the marketplaceinclude: (i) validation of single data-drivensegment solutions that are used to build anentire marketing plan; (ii) increased insightinto the changes encountered in the market-place; (iii) provision of a sound basis for fore-casting; (iv) the possibility of regularevaluation as to whether the segment(s) tar-geted should be held on to or switched; (v)temporary reduction of dimensionality dur-ing the segment assignment procedure; and(vi) applicability to typical multi-period datain tourism (non-panel format).

The aim of this chapter is to suggest aframework for tracking a posteriori marketsegment trends over years based on guest sur-veys conducted at different points in timebased on the same sampling method and thesame questionnaire.

The Tracking Framework

The tracking framework suggested is a step-wise process comprising seven steps.

1. Definition of the anchor period. Data-drivenmarket segmentation is no deterministic con-cept. The typically explorative nature of aposteriori segmentation has to be accountedfor in this concept by choosing an anchor

period for analysis. The anchor period isused as the starting point of the investiga-tion. Any period of survey data available canbe utilized: using the first period allows state-ments about the development that has takenplace to date, while using the last periodgives insight into how the present market sit-uation has developed. Clearly, if data for allperiods are available and the tracking frame-work is used for explorative ex post investiga-tion only, the definition of an anchor periodwould not be necessary. If, however, ongoingmonitoring is intended, the anchor periodchoice is inevitable. 2. Computation of a data-driven market segmenta-tion solution using data from the anchor period.Using the guest survey data set from theanchor period, a segment solution is derived.Any method that partitions the multi-dimen-sional data set in an appropriate manner canbe used at this stage. The result is a numberof market segments with each respondentassigned membership to one of the segments. 3. Characterization of market segments. Based onthe answers provided by segment members tothe segmentation base (Wedel andKamakura, 1998), the groups of tourists aredescribed in detail. Furthermore, relevantbackground variables (demographic, socio-economic, behavioural, etc.) are studied foreach segment. Using discriminant analysis atthis stage, for example, can help to validatethe existence of heterogeneous segments withregard to this background information. 4. Assignment of data from other periods to theanchor segment solution. The guest survey datafrom the remaining periods is matched withthe segment solution derived from theanchor year. This is achieved by extractingthe centroids from the anchor solution;these function as representants or proto-types for the segments. The answer patternsof respondents from the remaining periodsare then assigned to the closest correspond-ing prototypes. The result are frequency dis-tributions of segment assignments for eachperiod of time. 5. Testing of distribution changes. The distribu-tions of respondents over segments are com-pared over the various time periods applyingchi-squared tests based on contingency tables,including year and cluster membership infor-

246 S. Dolnicar


mation. Bonferroni correction of significancevalues is necessary if more that two periodsare studied. At this stage, it is possible todetermine whether there are any significanttrends in a posteriori segments over time. 6. (Testing of changes in background variables).In addition, qualitative changes can be stud-ied by investigating differences in back-ground variables for the same segmentsover years. 7. Validation of results. Validation of this step-wise procedure is of utmost importancebecause data-driven segmentation is anexploratory tool by itself and potentially ren-ders a million different solutions, one ofwhich is then chosen at random or by com-paring the usefulness of different solutions.By including multiple periods of time,another dimension of possible influence isincluded and this makes it even more dan-gerous to base the entire market structureinterpretation on one single run of analysis.Basically, repetition is a useful tool.Repetition can be conducted with differentnumbers of clusters, different algorithms ordifferent anchor years. By comparing solu-tions and time changes, a picture emergesfrom the exploratory approach that allowsconclusions to be made about the reliabilityof findings.

This stepwise tracking framework is flexiblein many respects: first, any kind of data canbe used that is appropriate for traditional aposteriori segmentation (multi-dimensionaldata on demographics, socio-economic,tourist behaviour, benefits sought, etc.).Second, the choice of the anchor year allowsa wide variety of explorative approaches (aswell as ongoing monitoring of a posteriorisegment development). Insights into segmentstructure development can thus be derivedfrom various perspectives. Third, any algo-rithm that results in a partitioning of answerpatterns of respondents can be applied, aslong as each respondent is assigned to a seg-ment deterministically. Finally, any back-ground variables can be chosen for validationand detailed description of the results.

In addition to the fundamental aim of thisprocedure, it might as well be used to deter-mine the optimal segment solution in thefirst place. For this purpose it is recom-

mended – as in the case of validation – to per-form the entire procedure a number of times,with changing anchor years, varying partition-ing algorithms and numbers of segments.Solutions with the highest stability would befavoured, unless distinct density segments canbe identified in the data (Dolnicar andLeisch, 2001).

Tracking Development of Activity-basedTourist Segments in Austria – an

Illustration

Austrian National Guest Survey data from thesummer seasons 1994 and 1997 were used toillustrate the tracking framework suggested.An activity-based segmentation is constructedwith the first survey used as anchor year. Thesample sizes amount to 7967 for the year 1994and 6604 for 1997. Respondents were askedto state which leisure activities they engage induring their vacation. In the data used, ‘1’indicates that the activity was undertakensometimes or often, whereas a ‘0’ indicateseither the fact that a respondent did notundertake that particular pastime or that heor she has not answered the question. Hence,the data set used is in binary format andincludes 14,571 respondents. Answers pro-vided by each of the respondents with regardto 22 vacation activities are used as the basisof segmentation. In addition, a number ofbackground variables are available, but theanalysis of background variables is omitted inthis empirical illustration, as inspection of thesegmentation base renders sufficiently illus-trative results.

Data from the first period of time is cho-sen as the anchor period. A self-organizingfeature map functions as partitioning algo-rithm (SOFM, Kohonen, 1984; for applica-tions in the field of a posteriori segmentationof guest surveys, see Mazanec, 1994, 1999;Dolnicar, 1997).

A map with six prototypes is used (threecolumns, two rows); starting points for theprototypes are chosen by drawing 100 pointsat random and picking the best solution asdetermined by the criterion of maximumbetween-segment variance and minimuminner-segment variance. The data set is pre-

Towards Data-driven Segmentation in Tourism 247


sented to the neural network 200 times forlearning purposes, with a decreasing amountof adaptation of both the prototype most sim-ilar to the answer pattern presented and theneighbouring prototypes in the grid. Afterthis learning phase, in which the SOFMadapts prototype values best to mirror thedata at hand both in terms of representationby six segments as well as with regard to topo-logical arrangement along the grid, each datavector is presented to the SOFM one moretime, each respondent being assigned to theprototype best representing his or her vaca-tion activity answer pattern.

The resulting segment profiles are pro-vided in Fig. 21.1 in an arrangement mirror-ing the SOFM grid. Each profile chartcharacterizes one of the segments thatemerged from the partitioning step. The barsindicate the average tendency of the segmentmembers to undertake the vacation activitieslisted and the line shows the total sampleaverage. Segment 1 could thus be describedas ‘culture tourists’. Members of this groupstate that they have engaged in cultural activi-ties of various kinds more often than the aver-age summer tourist in Austria, especially thelevel of ‘going to concerts’, ‘sightseeing’,‘going out in the evening’, ‘shopping’, ‘goingto the theatre’, ‘going to museums and exhi-bitions’ and ‘spending the evening at aHeurigen’ (this is a typical Viennese restau-rant). Segment 2 is less distinct, showinginterest in both cultural activities as well assports. Segment 3 is clearly sports-centredand segment 5 represents the typical relaxedsummer tourist who spends the days hikingand going for walks. The remaining two seg-ments are not interpreted, as it cannot be val-idated which proportion of the respondentsare ‘active in all respects’ or ‘not interested inany activities’, as opposed to being mereanswer tendencies that are concentrated insegments 4 and 6.

The contingency table in Table 21.1describes changes taking place over the 2years of the survey. The Pearson chi-squaredrenders a highly significant outcome at the99.9% significance level. The major trendswith regard to activity segment shifts from1994 to 1997 are: the increase of the sportssegment, an increase in hiking tourists, and a

dramatic decrease in respondents statingeither that they engage in all activities or havea positive answer tendency.

The results are validated by rerunning theentire process for ten segments with a differ-ent algorithm (topology representing net-works as introduced by Martinetz andSchulten, 1994) that functions in a similarway as self-organizing feature maps but doesnot force the prototypes into the predefinedgrid. Table 21.2 illustrates the relationbetween the two segmentation solutions. Theculture segment is split up in segments TRN 2and TRN 5 (cross-sections are pointed out inTable 21.2 in bold), the main differencebetween these segments being the fact thatno member of TRN 2 participates in orga-nized excursions, whereas all TRN 5 membersdo. Similarly, the sports-oriented segment issplit up in TRN 3 and TRN 6 with the majordiscriminating variable being the amount ofbiking done during the vacation. The hikingtourists remain very stable (matching TRN10), although the number of segments wasalmost doubled. The same is true for segment4: the potential negative answer tendency seg-ment is represented by prototype TRN 10.The remaining segments resulting from theten-segment solution are very difficult tointerpret. The cross-tabulation (significantchi-squared test at the 99.9% level) shows thatsegments 2 and 6 in the SOFM solution areboth split up among the four new segments.The ten-segment solution thus backs up seg-ments 1, 3, 4 and 5 from the SOFM solution,as well as supporting the fact that the remain-ing group of tourists are not easily segmentedin terms of vacation activities, as no stablerepresentation can be determined.

From the change tracking perspective overconsecutive survey years (see Table 21.3), theresults from the six-segment solution are sup-ported with regard to the increase in thesports activity segment, as well as in the hik-ing segment on the one hand and thedecrease of segment SOFM 6 (TRN 4, 7 and8) on the other.

All in all, validation of the initial six-segment solution by repeating the trackingprocess for a ten-segment solution with a dif-ferent partitioning algorithm revealed whichof the findings based on the SOFM analysis

248 S. Dolnicar



tenn

is biking

hors

ebac

k ridi

nggo

lf

swim

ming

going

to sp

as surfi

ng boat

ing

mou

ntain

eerin

g hiking

going

for w

alks

orga

nized

exc

ursio

ns

excu

rsion

sre

laxing

out in

the

even

ingsh

oppin

gco

ncer

ts

sight

seein

g

tyrole

an e

venin

gs

mus

eum

sth

eatre

heur

igen

1 aver

age

tenn

is biking

hors

ebac

k ridi

nggo

lf

swim

ming

going

to sp

as surfi

ng boat

ing

mou

ntain

eerin

g hiking

going

for w

alks

orga

nized

exc

ursio

ns

excu

rsion

sre

laxing

out in

the

even

ingsh

oppin

gco

ncer

ts

sight

seein

g

tyrole

an e

venin

gs

mus

eum

sth

eatre

heur

igen

4 aver

age

tenn

is biking

hors

ebac

k ridi

nggo

lf

swim

ming

going

to sp

as surfi

ng boat

ing

mou

ntain

eerin

g hiking

going

for w

alks

orga

nized

exc

ursio

ns

excu

rsion

sre

laxing

out in

the

even

ingsh

oppin

gco

ncer

ts

sight

seein

g

tyrole

an e

venin

gs

mus

eum

sth

eatre

heur

igen

tenn

is biking

hors

ebac

k ridi

nggo

lf

swim

ming

going

to sp

as surfi

ng boat

ing

mou

ntain

eerin

g hiking

going

for w

alks

orga

nized

exc

ursio

ns

excu

rsion

sre

laxing

out in

the

even

ingsh

oppin

gco

ncer

ts

sight

seein

g

tyrole

an e

venin

gs

mus

eum

sth

eatre

heur

igen

2 aver

age

5 aver

age

tenn

is biking

hors

ebac

k ridi

nggo

lf

swim

ming

going

to sp

as surfi

ng boat

ing

mou

ntain

eerin

g hiking

going

for w

alks

orga

nized

exc

ursio

ns

excu

rsion

sre

laxing

out in

the

even

ingsh

oppin

gco

ncer

ts

sight

seein

g

tyrole

an e

venin

gs

mus

eum

sth

eatre

heur

igen

tenn

is biking

hors

ebac

k ridi

nggo

lf

swim

ming

going

to sp

as surfi

ng boat

ing

mou

ntain

eerin

g hiking

going

for w

alks

orga

nized

exc

ursio

ns

excu

rsion

sre

laxing

out in

the

even

ingsh

oppin

gco

ncer

ts

sight

seein

g

tyrole

an e

venin

gs

mus

eum

sth

eatre

heur

igen

3 aver

age

6 aver

age

Per cent

0102030405060708090100

Per cent

0102030405060708090100

0102030405060708090100 010203040506070809010

0

0102030405060708090100 010203040506070809010

0

Fig.

21.

1.Se

gmen

t pro

files

bas

ed o

n a

six-

prot

otyp

e SO

FM s

olut

ion

for

1994

.


can be built on for planning marketing actionand which findings have to be used with cau-tion. The ‘culture segment’, the ‘sports seg-ment’, the ‘hiking segment’ and the groupengaging in very few vacation activities (orthe negative answer tendencies group) can beaccepted as relatively stable, and the samemarket trends concerning these groups oftourists result from both investigations. Theremaining segments are not identified as sta-ble and therefore should not be chosen astarget segments without further investigation.

Conclusions

This chapter has proposed a tracking frame-work for the tracing of trends in a posteriorisegments over time. The stepwise frameworkincludes the following stages: (i) definitionof the anchor period; (ii) computation of adata-driven market segmentation solutionusing data from the anchor period; (iii) char-acterization of market segments; (iv) assign-ment of data from other periods to theanchor segment solution; (v) testing of distri-

250 S. Dolnicar

Table 21.1. Contingency table for segment size comparison 1994 and 1997 (SOFMbased).

Segment 1994 1997 Total

1 (Culture) Number of respondents 1,672 1,337 3,009% Within the year 21% 20% 21%

2 (Culture and sports) Number of respondents 1,321 1,116 2,437% Within the year 17% 17% 17%

3 (Sports) Number of respondents 1,499 1,474 2,973% Within the year 19% 22% 20%

4 (Nothing, answer tend.) Number of respondents 1,034 848 1,882% Within the year 13% 13% 13%

5 (Hiking) Number of respondents 1,241 1,233 2,474% Within the year 16% 19% 17%

6 (All, answer tend.) Number of respondents 1,200 596 1,796% Within the year 15% 9% 12%

Total Number of respondents 7,967 6,604 14,571% Within the year 100% 100% 100%

Table 21.2. Cross-tabulation of the six-segment SOFM and the ten-segment TRN solution.

SOFM with sixTRN solution with ten segments

segments 1 2 3 4 5 6 7 8 9 10 Total

1 No. of members 893 16 522 3 1 1,435% SOFM 62% 1% 36%2 No. of members 874 10 131 8 3 1,311 151 8 1 2,497% SOFM 35% 5% 53% 6%3 No. of members 1 2,116 51 4 1,495 69 12 5 3,753% SOFM 56% 1% 40% 2%4 No. of members 20 249 22 4 106 19 91 1,724 2,235% SOFM 1% 11% 1% 5% 1% 4% 77%5 No. of members 36 69 1 412 31 60 9 2,355 88 3,061% SOFM 1% 2% 13% 1% 2% 77% 3%6 No. of members 244 26 33 574 18 80 246 531 1,752% SOFM 14% 1% 2% 33% 1% 5% 14% 30%Total no. of

members 1,155 1,018 2,399 1,206 587 1,747 1,576 761 2,466 1,818 14,733


bution changes; (vi) testing of changes inbackground variables; and (vii) validation ofresults, with the last step being of centralimportance in reducing the influence of ran-dom factors in the final result. The frame-work is flexible regarding methodologicalapproaches preferred at each step.

In order to apply the tracking frameworkthe procedure requires multiple-period guestsurvey data based on an identical question-naire and sampling strategy.

The advantages of being able to track aposteriori segment trends in the marketplaceinclude: (i) validation of single data-drivensegment solutions on which an entire market-ing plan can be built; (ii) increased insightinto the changes encountered in the market-place; (iii) provision of a sound basis for fore-casting; (iv) the possibility of regularevaluation as to whether the segment(s) tar-geted should be maintained or replaced; (v)only temporary reduction of dimensionalityduring the segment assignment procedure

(no compression of the item information);and (vi) applicability to typical multi-perioddata in tourism. (It is not necessary for thesame individuals to be questioned. Such arequirement would be difficult to achieve inthe case of touristic guest surveys.)

The main limitation of the concept is theinsecurity arising from sampling over consec-utive periods of time. It cannot automaticallybe assumed that multiple surveys based onrepresentative samples are sufficient toexclude intervening variable effects that maydistort the results. This problem is especiallycrucial in the case of surveys that rely stronglyon weighting cases due to sampling restric-tions, but this limitation clearly is not causedby the tracking framework proposed butaffects any analysis of data sets derived in typi-cal guest survey manner over multiple peri-ods of time.

Further work is needed to extend thedata-driven segment tracking tool in a three-way data situation, thus building upon the


Table 21.3. Contingency table for segment size comparison 1994 and1997 (TRN based).

Segment 1994 1997 Total

1 Number of respondents 622 556 1,178% Within the year 8% 8% 8%







8 Number of respondents 523 306 829% Within the year 7% 5% 6%



Total Number of respondents 7,967 6,604 14,571% Within the year 100% 100% 100%


perceptions-based market segmentation con-cept as first introduced by Dolnicar et al.(1999) and comprehensively described inMazanec and Strasser (2000). The relevanceof this extension is founded on the increas-ing importance of brand and destinationimage studies. Extending the tracking tool insuch a way would enable simultaneous trendtracking of a posteriori segmentation, posi-tioning and competition in the marketplace.

Acknowledgements

This study was partially conducted duringSara Dolnicar’s appointment as assistant pro-fessor at the Institute for Tourism and LeisureStudies at the Vienna University ofEconomics and Business Administration andsupported by the FWF grant 010 (AdaptiveInformation Systems and Modeling inEconomics and Management Science).

252 S. Dolnicar

References

Baumann, R. (2000) Marktsegmentierung in den Sozial- und Wirtschaftswissenschaften: eine Metaanalyseder Zielsetzungen und Zugänge. Diploma thesis at Vienna University of Economics and ManagementScience, Vienna.

Dolnicar, S. (1997) Erstellung einer psychographischen Taxonomie von Osterreich-Urlaubern unter Anwendung neu-ronaler Netzwerkverfahren. Service Fachverlag, Vienna.

Dolnicar, S. (2002) Review of data-driven market segmentation in tourism. Journal of Travel and TourismMarketing 12, 1–22.

Dolnicar, S. and Leisch, F. (2001) Knowing what you get – a conceptual clustering framework for increasedtransparency of market segmentation studies. Abstract Proceedings of the Marketing Science Conference inWiesbaden, INFORMS, Gelnhausen, pp. 88–89.

Dolnicar, S., Grabler, K. and Mazanec, J.A. (1999) A tale of three cities: perceptual charting for analyzingdestination images. In: Woodside, A.G., Crouch, G.I., Mazanec, J.A., Oppermann, M. and Sakai, M.Y.(eds) Consumer Psychology of Tourism, Hospitality and Leisure. CAB International, Wallingford, UK,pp. 39–62.

Ketchen, D.J. Jr and Shook, C.L. (1996) The application of cluster analysis in strategic managementresearch: an analysis and critique. Strategic Management Journal 17, 441–458.

Kohonen, T. (1984) Self Organisation and Associative Memory. Springer-Verlag, New York.Lilien, G.L. and Rangaswamy, A. (1998) Marketing Engineering. Addison-Wesley, Reading.Martinetz, Th. and Schulten, K. (1994) Topology representing networks. Neural Networks 7, 507–522. Mazanec, J. (1994) Image measurement with self-organizing maps: a tentative application to Austrian tour

operators. Revue de Tourisme 49, 9–18.Mazanec, J. (1999) Simultaneous positioning and segmentation analysis with topologically ordered feature

maps: a tour operator example. Journal of Retailing and Consumer Services (Special Issue on MarketingApplications of Neural Networks) 6, 219–235.

Mazanec, J. (2000) Market segmentation. In: Jafari, J. (ed.) Encyclopedia of Tourism. Routledge, London.Mazanec, J. and Strasser, H. (2000) A Nonparametric Approach to Perceptions-based Market Segmentation:

Foundations. Springer, Berlin.Middleton, V.T.C. (1988) Marketing in Travel and Tourism. Heinemann, Oxford. Myers, J.H. (1996) Segmentation and Positioning for Strategic Marketing Decisions. American Marketing

Association, Chicago, Illinois.Punj, G. and Stewart, D.W. (1983) Cluster analysis in marketing research: review and suggestions for appli-

cation. Journal of Marketing Research 20, 134–148.Smith, S.L.J. (1995) Tourism Analysis. Longman, Harlow. Wedel, M. and Kamakura, W. (1998) Market Segmentation – Conceptual and Methodological Foundations. Kluwer

Academic Publishers, Boston, Massachusetts.


Chapter twenty-twoSustainable Tourism and Stakeholder Groups: a Case Studyof Colorado Ski Resort Communities

Richard R. PerdueLeeds School of Business, University of Colorado at Boulder, Boulder, CO 80309-0419, USA

Abstract

The purpose of this chapter is to illustrate the application of stakeholder theory in the ski resort industryof Colorado. Sustainable development is increasingly a core concept of tourism management. Stakeholdertheory is proposed as a means of assuring the sustainable development goal of intra-generational equity.The services marketing triangle is extended to identify three critical resort community populations: guests,resort employees and host community residents. Using a combination of quantitative and qualitativemethodologies, eight stakeholder groups within these three populations are identified and described.

Introduction

Over the past two decades, sustainable devel-opment has become a core concept oftourism planning (Pigram, 1990; Hunter,1997; Smith, 2001). A substantial body ofresearch has evolved focusing on the defini-tion of sustainable tourism (Garrod and Fyall,1998), determining dimensions of sustainabil-ity (Hunter, 1997), testing alternative plan-ning methodologies (Berry and Ladkin,1997; Ahn et al., 2002), and identifying mea-surement metrics (Shane and Graedel, 2000;McCool et al., 2001; Miller, 2001).

While there are still many unansweredquestions, equity is generally recognized as acentral goal of sustainable tourism (Ryan,2002; WBCSD, 2002). Beginning with the

Brundtland Report (WCED, 1987) and theEarth Summit Local Agenda 21 (Jackson andMorpeth, 1999), both inter- and intra-genera-tional equity are essential goals of sustainabletourism development. Inter-generationalequity focuses on resource conservation forfuture generations. Intra-generational equityemphasizes fairness to various population sec-tors both in access to the natural and culturalresources that support tourism and in voiceconcerning resource allocation decisions.

Stakeholder theory has been proposed asa theoretical framework for assessing intra-generational equity in tourism planning(Sautter and Leisen, 1999). Essentially, stake-holder theory postulates: (i) that a firm hasrelationships with many constituent groups(stakeholders) that affect and are affected by



its decisions; (ii) that the interests of allstakeholders have intrinsic value and no setof interests is assumed to dominate the oth-ers, particularly over time; and (iii) manager-ial decision making is enhanced byunderstanding and, to the extent possible,sensitivity to the various stakeholders(Freeman, 1984; Friedman and Miles, 2002).While divergent streams of stakeholder the-ory exist, the basic premise is the importanceof identifying key stakeholder groups andunderstanding their key concerns (Jones andWicks, 1999).

The purpose of this chapter is to illustratethe application of stakeholder theory in theski resort industry of Colorado. With the con-tinuing expansion of the resorts, conflictsover resource decisions are increasingly hos-tile and subject to extensive litigation withhigh associated time and financial costs.Effectively working with concerned groups iscritical to virtually any development or expan-sion initiative. Stakeholder theory provides aframework both for identifying the variousgroups and understanding their concerns.

The Evolution of Ski Resorts in Colorado

Snow skiing has always been in the lifebloodof Colorado, dating from the mid-1800s whenminers, mail carriers and ministers used their‘Norwegian snowshoes’ to travel throughout

the mountains (Fay, 2000). Ski races wereorganized by clubs in Irwin and Crested Butteas early as 1886. However, it was not until the1930s and 1940s that the earliest ski resortswere established in Winter Park and Aspen.The heyday of Colorado destination ski resortdevelopment was the 1960s with the develop-ment of Vail, Crested Butte, Breckenridge,Snowmass, Steamboat Springs, Purgatory,Sunlight, Powderhorn and Keystone Resorts(Fay, 2000).

As reflected in Fig. 22.1, the Colorado skiindustry has matured over the past decade.After enjoying robust growth between the1968/69 and the 1992/93 seasons, total skierdays have been relatively flat since 1993 atslightly over 10 million per year. The varia-tions between 1993 and 2001 are primarily afunction of snow conditions. The effects ofthe recessions of 1980/81, 1990/91, and2001/02 are evident, as are the effects of thesevere, pre-snowmaking drought of the1976/77 season and the poor snow condi-tions of the 1998/99 and 1999/2000 seasons.A variety of factors, including an aging USpopulation and the rapid growth of both thecruise industry and beach vacation destina-tions in the southern USA and Mexico, areviewed as contributing to the stagnantdemand for skiing.

Thus, Colorado ski resorts are increasinglycompeting for share in a stagnant market. Aprimary tactic in this competition has been

254 R.R. Perdue

0

2

4

6

8

10

12

Season

Ski

er d

ays

(mill

ions

)

68/69 01/02

Fig. 22.1. Colorado skier days.


product improvements in the form of terrainexpansions and lift capacity improvements,both of which are very expensive. Replacing astandard double chairlift with a high speeddetachable quad lift is at least US$6–8 mil-lion. A gondola lift costs in the range of$10–12 million. Even when excluding thehighly variable litigation costs of gaining thedevelopment permits, the costs of terrainexpansion are equally impressive; the BlueSky Basin expansion at Vail Resorts is esti-mated to have cost $18 million. Further,today’s destination skier also expects a com-plete range of restaurant facilities on themountains, other winter recreation facilitiessuch as sledding and ice skating, and otheramenities, such as warming huts and businesscentres. The Two Elk Restaurant complex onVail Mountain is a $12-million facility. Overthe last decade, more than $1 billion hasbeen invested in mountain expansions andimprovements at existing Colorado ski resorts(Lipsher, 2002).

Additionally, it is important to recognizethat real estate development/sales and prop-erty management are critical sources ofincome for most Colorado ski resorts. The skiareas are predominantly located on publiclyowned lands, primarily US Forest Servicelands, but the base villages are privatelyowned. The development and managementof vacation homes, condominiums and timeshares is a major source of resort income.Most of this real estate is sold with the agree-ment that the resort will manage it as rentalproperty when the owner is not present with atypical 50/50 split of the rental income. Thekey sales presentation tool is an impliedpromise, based on historical data, that theowner’s share of the rental income will pay allownership costs, including both mortgagepayments and maintenance fees. Over thepast decade, both the number of vacationproperties at most Colorado resorts and thecost of these properties have more than dou-bled, meaning that the resorts must dramati-cally increase rental income in order to fulfillthis promise.

This combination of very high infrastruc-ture costs and need for rental income funda-mentally means that the resorts cannot besuccessful as seasonal businesses. Consequently,

they have been actively developing both sum-mer recreation facilities and products, such asgolf courses, mountain biking, and music festi-vals, and non-seasonal facilities such as confer-ence centres and meeting facilities. Eliminatingthe seasonal variations in resort demand is acritical strategic marketing goal.

In summary, over the past 10–15 years, skiresorts in Colorado have been aggressivelyexpanding ski terrain, ski facilities, restau-rants, mountain amenities, vacation realestate, non-skiing winter and summer recre-ation areas, and conference/meeting facili-ties. Each development must be approvedand permitted by, at a minimum, the localand county government planning commis-sions. Additionally, most mountain expan-sions also require federal governmentpermits. The time and litigation costs ofobtaining these permits have grown exponen-tially. Working with key stakeholder groupsboth formally in public meetings and infor-mally by other means is a necessary compo-nent of attaining these permits and reducingthe associated costs.

The Services Marketing Triangle

The services marketing triangle was used asa framework for identifying the stakeholdergroups. Service quality is widely recognizedas a critical determinant of success in thetourism and leisure industries (Kotler et al.,1999; Perdue, 2000). In the Colorado skiresort industry, at least 80% of the guestsare a direct result of service quality, eitheras satisfied repeat guests or first time gueststhere as a result of a word of mouth recom-mendation. As a tool for understandingquality, the services marketing triangle(Zeithaml and Bitner, 2000) postulates thatquality is a function of the interactionsbetween a services firm and two populations– service consumers and service employees.As reflected both by the wide range of hostcommunity resident attitude studies and bythe political process of developmentapproval, adapting this concept to a sustain-able tourism planning context necessitatesadding a third population, host communityresidents. Within the Colorado resort com-

Sustainable Tourism and Stakeholder Groups 255


munities, each of these populations isextremely and increasingly diverse, implyingthe need to identify key stakeholder groupswithin each population.

Historically, tourism service qualityresearch has focused on measuring consumerperceptions of quality and its impact on con-sumer satisfaction and revisit intentions (Fickand Brent Ritchie, 1991; Yuksel and Yuksel,2001). Market segmentation has been a majorpart of this research (Noe, 1999). Clearly,there is enormous diversity in the characteris-tics and behaviours of leisure participants(Oliver, 1997). A broad portfolio of measuresand techniques for segmenting this diversepopulation into more homogeneous groupshas subsequently evolved (Driver andJohnston, 2001) and is applied in this chapteras the foundation for identifying the key con-sumer stakeholder groups.

Quality employees are the raison d’être ofservice quality in resort settings (Spiselman,1995; Roos, 2002). Yet, for a variety of rea-sons, Colorado resorts are finding it increas-ingly difficult to attract and retain goodemployees (Ledgerwood et al., 1998).Consequently, the resorts have developed avariety of recruitment strategies aimed atboth traditional and non-traditionalemployees. The result is an increasinglydiverse employee population. Employeestakeholder groups are identified as ameans of understanding this diversity ofemployees, including not only their jobmotivations, expectations and behaviours(Bowen et al., 2000; Brown and Dev, 2000)but also their reactions to alternative devel-opment scenarios.

At least four critical interactions betweenresorts and local residents have been identi-fied: (i) local residents are the primarysource of employees; (ii) they commonlyinteract with resort guests in the host com-munity and at the resort; (iii) they can initi-ate anti-tourism behaviours that quickly anddramatically damage a resort’s image; and(iv) through the political process, they canimpact public policy toward tourism develop-ment (Perdue et al., 1999). While there is asignificant body of research on host commu-nity residents and their attitudes towardtourists and tourism development (Anderek

and Vogt, 2000), this research has not beenadequately extended to address the increas-ing diversity of resort community residents.Over the past decade, many Colorado resortcommunities have experienced extraordinar-ily high population growth. A diverse mix-ture of people has migrated to the resortcommunities for a variety of different rea-sons, many of which are not related to workor employment. Again, identifying stake-holder groups is a means of understandingthis diversity of residents and their attitudestoward sustainable tourism.

Study Methodology

This chapter is a summary of research con-ducted over the past 10 years. For the skierpopulations, the associated research includesyearly on-site skier surveys and focus groupstudies, a panel study of the primary localskier market – The Colorado Front Rangepopulation, which includes the Denver,Boulder, Colorado Springs, and Fort Collinsmetropolitan areas, and three major studiesof the destination skier market conductedusing mail or Internet survey methodologies.The employee studies include an annualemployee opinion survey that has been con-ducted at four major resorts for the past7 years, focus groups and qualitative inter-views. The host community resident studiesinclude surveys conducted in five resort com-munities. Additionally, there have been ongo-ing participant observation studies, extendedinterviews with resort managers and commu-nity leaders, and continual analysis of bothindustry and news media articles. Finally,extensive secondary analyses have been con-ducted of resort operating data, includingfinancial reports, consumer satisfaction andcomplaint data, and guest history databases.

Colorado Resort Community StakeholderGroups

The central propositions of this chapter arethat: (i) sustainable tourism development inColorado resort communities is predicated

256 R.R. Perdue


on understanding the various stakeholdergroups who are likely to become involved inthe public policy and development permit-ting process; and (ii) stakeholder groups existwithin three critical populations: resortguests, resort employees and host communityresidents. Over the past decade, each of thesepopulations has become increasingly diverse,comprising substantively different values,opinions and behaviours toward resort devel-opment and expansion. The followingdescribes eight stakeholder groups: two resortguest groupings, three resort employeegroups and three host community groups.Recognize that this grouping is not compre-hensive; its purpose is to illustrate the diver-sity of these critical populations.

Employee groups

Traditional ski bums

Probably, the most recognized of theemployee stakeholder groups is the tradi-tional ski bum. They are young, typicallymale, from a reasonably affluent background,and are relatively well educated. Many ofthem are taking a semester or year off fromtheir university education to work at the skiresort. Many others have just finished theiruniversity education and are working for aseason at a ski resort before they ‘get serious’about their careers and enter the ‘real world’.Importantly, ‘ski bums’ often feel over-quali-fied for the jobs they are being asked to per-form. Many jobs, such as lift operations, tendto be repetitive and can be very boring.Further, because of their affluent back-ground, they seldom have experience work-ing in front-line service occupations. As aresult, many of them resent the nature oftheir jobs.

The ski bum loves to ski. They work at aski resort so that they can ski. Most of themrank the opportunity to ski as critical to theirlife satisfaction. Often they ski as many as 100days a winter, which equates to 2 out of every3 days. Comparatively, the ski bum’s job isrelatively unimportant to them. Many changejobs several times over the course of the skiseason, which is easy to do because of the

high demand for employees in most resortcommunities. They are apt to catch the ‘pow-der flu’, which is failure to show up for workon days with fresh powder snow on themountains.

The traditional ski bum is a very frustrat-ing employee. They are not motivated bytheir jobs, their salaries or career aspirations.They are there to ski. Work is very secondary.If they do not like their current job, it is veryeasy to find another one quickly. Many havehigh absenteeism levels. Resort managershave to work extraordinarily hard to makethe job fun and exciting to this group.Further, they have to make the employee feelinvolved in the resort.

Poor service quality can be a serious prob-lem. Traditional ski bums often developderogative terms for resort visitors anddevelop an attitude of superiority, primarilybased on their superior ski talent.Additionally, because of their resentmenttoward their jobs, they frequently resent thespecial requests and questions of resort guests.

Interestingly, this group further dividesinto three groups. Most return to theirhomes, their education and their plannedcareers after the ski season. However, somestay in the resort communities and, over time,are promoted into management careers withthe resorts. These individuals often developvery strong attachments to the resort commu-nities and their summer activities. A commonsaying in this group is that ‘I came for thewinter, but stayed for the summers’.

A third group has evolved at upscaleresorts such as Aspen and Vail – the profes-sional ski instructor. On a typical day at theseresorts, a relatively large number of skierspurchase private lessons, frequently as ameans of avoiding lift lines. The professionalski instructor has evolved to serve this mar-ket. This group is highly evolved, is very closeknit, has its own language and a very strongsocial structure based on seniority and skiability. This group develops a very strongsense of ‘insiders’. Since gratuity income isvery important, they develop a strong sense ofservice quality for their clients. However, theyalso develop an equally strong sense of dis-dain for skiers who do not purchase lessons.



The new ski bum

A second group of ski bums has emergedover the past 10 years. This group is almostthe opposite of the more traditional ski bum.They are older, often over 50. Many areretirees, having taken early retirement fromcorporations or from military careers. Theyare reasonably affluent in that they haveincome from pensions and retirement funds.

They love to ski, but more importantly,they love to live in the resort towns. Theybecome very attached to the resort commu-nity and are often active in local programmesand activities. Many of them enjoy the sum-mers as much or more than the winters.

The new ski bum works in the resorts forthe social opportunities of meeting otheremployees and guests, and because it givesthem a sense of belonging to the ski commu-nity. While their jobs are important forincome, they are even more important for thesocial and belonging opportunities.

They tend to be excellent employees. Theyare reliable, consistent, friendly and far lessderogatory in their attitudes toward resortguests. Obviously, they are more mature andmore experienced.

However, the new ski bum is also very con-cerned about the effects that resorts have onthe environment and the local communities.Quality of life and the cost of living in theresort communities are very important issuesto these individuals. They are likely actively tooppose resort development and expansion ifthey feel it will cause declines in local qualityof life or increases in the cost of living.

Migrants

Getting and keeping quality employees is per-haps the greatest resort management chal-lenge in Colorado, particularly for manyback-of-the-house jobs such as housekeepingand maintenance. Over the past two decades,there has been a dramatic growth in the num-ber of migrant workers. Initially this groupwas comprised almost exclusively of peoplefrom Mexico. More recently, there has been avery strong surge in Eastern Europeans andAustralians.

The Hispanic migrant workers enjoy a fas-cinating leisure lifestyle. They work extremelyhard for 6 months and then spend the sum-mers not working, enjoying their leisure backin Mexico. The resorts help them to attain a6-month work visa. During those 6 months,they try to work as many hours as possible,often working several different jobs and aver-aging in excess of 80 h of work per week.

One of the potentially significant prob-lems with the migrant employees is thenature of their work visa. The visa is actuallygranted to the resort. If the employee quits oris terminated from their resort job, the visa isrevoked and the migrant has to immediatelyreturn to his/her home country. This createsan opportunity for employee abuse and mis-treatment.

The resorts feel that the migrant Hispanicworkers are excellent employees. They showup for work every day. They will accept jobsthat many traditional ski bums will not do.They want to work as many hours as possibleand readily accept overtime assignments.Most do not ski and, consequently, never getthe ‘powder flu’. Contrary to the ski bums,they appreciate their jobs and rarely com-plain or cause problems with resort guests.However, they obviously require the resorts todevelop both language and cultural sensitivitymanagement skills.

Additionally, the resort communities arerequired to acclimate to the Hispanic cultureand presence, greatly increasing pressure andcosts on such systems as housing, education,medical services and public safety. Of particu-lar note and concern to many resort commu-nity leaders and residents is the willingness ofthe Hispanic employees to live in far greaternumbers in apartments, often having eight toten people living in a simple two-bedroomapartment.

Because of the high cost of living in resortcommunities, it is not uncommon for themigrant employees to live in other communi-ties and commute either by car or by resort-provided busing. There are a number ofexamples in Colorado where the employeesare living in communities in adjacent‘employee-bedroom’ counties. Because bothlocal sales taxes and real estate taxes arelevied at the county level, the adjacent coun-

258 R.R. Perdue


ties do not benefit from the resort’s commer-cial success. Yet, these adjacent counties haveto develop and pay for social services pro-vided to the migrants, including bilingualeducation systems, medical services and socialwelfare services. This often puts an enormousburden on the adjacent county public ser-vices and tax base.

Host community resident groups

Trust fund babies

An enormous transfer of wealth is currentlyoccurring in the USA. The ‘depression’ gen-eration, born in the 1920s and 1930s, isdying. That generation is renowned for itswork-oriented lifestyle and for their propen-sity to save money. The next generation, thebaby boomers, is inheriting this money. Thebaby boomer lifestyle tends to be character-ized by a propensity to spend, often beyondtheir income as evidenced by high credit carddebt, lack of home equity and low savingsrate. The baby boomers are much more ori-ented to their current life satisfaction and arenot willing to defer leisure in favour of work.Thus, a very large sum of money is transfer-ring from a work/savings-oriented generationto a leisure/spending-oriented generation.Estimates of the size of this wealth transferover the next 20 years are as high as US$3trillion. As much as 5% of this money is beingspent on vacation homes in and aroundresorts throughout the USA, leading to a mas-sive build-up of resort condominiums andhomes. Virtually every resort community inColorado has experienced massive develop-ment and redevelopment of vacation proper-ties over the past decade.

This is leading to dramatic increases inprices. A single condominium in BeaverCreek sold recently for a record $7.8 million.Another sold in Aspen for $7.5 million.Numerous homes and estates have sold forprices in excess of $40 million. The local resi-dents who own property are benefiting fromthis price inflation. Those that rent or arewishing to buy property are being pushed outof the resort communities. For example, arecent study in Aspen, Colorado, concluded

that a typical local employee would have towork 27 full-time jobs in order to afford ahome in Aspen.

While many of these resort condominiumsand homes are being purchased as vacationproperties, there has also been a dramaticincrease in the number of very wealthy indi-viduals either permanently or semi-perma-nently living in the resort communities. Thisgroup has been entitled the ‘trust fund babies’in that their primary source of income isinheritance, in many cases leading to a popu-lation that does not need to work for income.The trust fund babies are a very wealthy,mostly Caucasian, middle-age group that isvery leisure oriented. They participate in ski-ing, but also expect and are patrons of highculture, entertainment and arts activities.

Importantly, the trust fund baby oftenviews the resort community almost as a pri-vate country club. They do not necessarilywant resort expansion or large numbers ofskiers. They live the ultimate leisure lifestyleand want a system that responds to theirevery wish and whim. Often, they lack toler-ance for less affluent groups, particularly theresort migrant employees. Conversely, theyare often ridiculed by those other popula-tions for conspicuous consumption, particu-larly for the size and opulence of theirhomes. The conversion of traditional agricul-tural and forested lands into the ‘trust fundbaby’ palatial estates is a significant politicalissue not only in the resort communities, butfor the entire State of Colorado.

Consultants

The ‘consultants’ are another group that hasmigrated into the resort communities inrecent years. This group is affluent, middleaged and highly educated. They work veryhard at jobs that are a combination of homeoffices, connected via the Internet to aremote corporation, and extensive travel totemporary job sites located throughout theworld. For work, they need a combination ofgood air access and broadband Internet ser-vices. They can get these services in the resortcommunities. Additionally, the resort commu-nities also provide the desired leisure lifestyleand quality of life. While the consultants work



many hours, those hours tend to be bundled.They also tend to have bundled leisure timewhen they participate not only in the resortski and summer activities, but also in theentertainment and related activities availablein resort communities.

In the USA, real estate taxes are a primarysource of funding for social welfare servicessuch as education, public safety, parks andrecreation, and medical services. ManyColorado resort communities have a verylarge vacation-home real estate base. Forexample, in Routt County (SteamboatSprings), Eagle County (Vail, Beaver Creek),Summit County (Keystone, Breckenridge,Copper Mountain), Pitkin County (Aspen,Snowmass) and San Miquel County(Telluride), the majority of the private prop-erty is owned by non-residents; frequentlythese properties are only owner-occupied afew weeks of the year. This creates a very largetax base for a relatively small resident popula-tion. As a result, many resort communitieshave extraordinarily excellent schools, med-ical services, recreation facilities and services,arts and entertainment, and public safety pro-grammes. Because of the mobility of theirjobs, the consultant can easily live in theresort communities and benefit from thisquality of life.

As with the trust fund babies, consultantsdo not depend on resort success for theirincome. Their primary focus is maintainingand enhancing the local quality of life, whichis not necessarily based on the local economy,particularly in the short term. The resortsneed to be able to demonstrate how theirexpansion and operations benefit the localquality of life to gain the support of this group.

Entrepreneurs

A recent study conducted in the area surround-ing Yellowstone National Park found that 40%of the local business owners had visited thearea first as a tourist, loved the area, and subse-quently moved to the region and either startedor purchased a business (Snepenger et al.,1995). The ski resort communities of Coloradohave also experienced this phenomenon.There is a growing population segment of‘entrepreneurs’ who have moved to the region

and started businesses because of their desireto live in an area that suits their leisure lifestyle.They are middle-aged and relatively affluent,but need their companies to be successful inorder for them to be able to afford to live inthe resort communities.

This group experiences very high frustra-tion levels. Because of the shortage ofemployees, they tend to work many hours,often to the detriment of their leisure partici-pation and the leisure lifestyle that originallyattracted them to the area. Many entrepre-neurs view the resorts both as business part-ners and as competitors. Because most oftheir businesses focus in some way ontourism, the resort’s success in attractingtourists to the area directly impacts their ownsuccess. However, the resorts are aggressivelytrying to retain a greater revenue share of vis-itor expenditures by offering an increasinglybroad range of products and services.Further, the resorts compete with the entre-preneurs for employees. Consequently, theentrepreneurs support resort expansion inany business category other than their own.They are adamant in their wish for the largeresort corporations to build and financeemployee housing, but are commonly unwill-ing to provide or subsidize housing for theirown employees.

Guest groups

As would be expected, there is enormousdiversity within the resort guest population;diversity that could be described at several lev-els of specificity. The following focuses only atthe very first level, the key differences betweenthe ‘local skiers’ and the ‘destination skiers’.

Local skiers

The local skiers are generally young,Caucasian, relatively affluent and well edu-cated. They live primarily in the metropolitancommunities on the Front Range of Coloradoand commute by personal car to the ski areas,primarily for day or weekend trips. Theseindividuals view skiing as a part of their nor-mal recreational activities as opposed to adeserved vacation experience.

260 R.R. Perdue


The local skiers participation is heavily dri-ven by value. They tend to be very price sensi-tive and are trying to get the best valuepossible, so that they can ski as often as possi-ble. In response, many of the resorts offerlocal skiers very heavily discounted seasonpasses; while the resort may charge as muchas $70 for a daily lift ticket, unlimited seasonpasses are available to local skiers for less than$300. Still, the local skiers commonly feel thatthe resort is over-charging for other products,particularly food, beverages and equipment.Many see the resorts as being too profit-ori-ented, too focused on the more lucrative des-tination skiers, and not sensitive to theirneeds.

The local skiers commonly developderogatory attitudes and opinions of the des-tination skiers. In a paradoxical sense, theylove to be tourists, but generally dislike othertourists. It is not uncommon for conflict andarguments to occur between the local anddestination skiers. The destination skiers arefrequently relatively inexperienced andunskilled. Consequently, the local skiers seethem both as a potential hazard and as a sub-ject of ridicule.

The greatest frustration for the local skiersis the heavy traffic on Interstate 70, the pri-mary access between the metropolitan FrontRange and the ski resorts. Many feel that theresorts should help pay for proposed multi-billion dollar improvements of the Interstate,but would adamantly oppose lift ticket priceincreases to pay for such improvements.

Local skiers generally oppose resort expan-sion both because of traffic concerns andenvironmental protection. Although many ofthem own resort real estate, the ongoing rapidexpansion of the resort communities is amajor environmental concern, particularlythe development of large ‘trophy home’estates owned by wealthy non-residents.

Destination skiers

Other than their geographical residence, thedestination skiers are demographically similarto the local skiers. They also tend to be youngto middle-aged, Caucasian and relatively welleducated. While their mean income is signifi-cantly greater than the local skiers incomes, the

median incomes are generally similar.Motivationally, they are fundamentally differ-ent. Instead of viewing skiing as a normalweekly recreational activity, they are on a skivacation and want a memorable experience.The common attitude is ‘I work hard, I’m onvacation, and I deserve the best’. As a result,the destination skiers tend to be less price sen-sitive, but can still be angered by the discrep-ancy between local and destination skier prices.

The destination skier is sensitive to crowd-ing both in the ski lift lines and in the resortcafeterias. The local skiers commonly bring atleast part of their lunch to the resort withthem and purchase part from the resort cafe-terias. The destination skier resents such‘brown baggers’ when the cafeterias arecrowded and lack seating, particularly giventhat they commonly pay very high prices fortheir food and beverages.

Interestingly, the destination skier bifur-cates into two sub-groups. One sub-group isvery novelty-seeking and has a ‘belt notchmentality’. They are trying to ski as many dif-ferent resorts as possible, which creates a sig-nificant competitive advantage for resortareas such as Summit County Colorado,where there are seven major ski areas withinan hour’s drive.

Conversely, the other destination skier sub-group is very loyal to a particular resort, oftenvisiting it both in the winter and summer sea-sons. Many of this group purchase resort con-dominiums and real estate that the resortsubsequently manages for them as a rentalproperty. This transition from ‘loyal guest’ to‘owner’ results in dramatically higher servicequality expectations and greater concern forsustainable tourism development.

Conclusions and Implications

Thus, there is extreme diversity within thestakeholder groups in Colorado resort com-munities, impacting virtually all aspects ofresort development, management and mar-keting. Subsidized employee housing devel-opments are but one example of thisdiversity of opinion and attitudes. As previ-ously mentioned, finding and keeping qual-ity employees is extremely difficult in the



resort communities. The high cost of livingis recognized as a major contributing cause.Consequently, there have been a number ofsubsidized housing initiatives proposed byvarious governments and organizations inthe resort communities. Each stakeholdergroup has a different reaction to these pro-posals. Trust fund babies and consultants aregenerally opposed due to fears of its impacton their lifestyles. Entrepreneurs are hugelyin favour, but expect the large resorts tofinance and develop the housing. The tradi-tional ski bums and migrant workers favourhousing development, but are concernedover housing restrictions, particularly themigrant workers’ concerns over occupancyrestrictions. The new ski bums are lessfavourable due to fear of its potential impacton their own housing investment; NIMBY(not in my backyard) is a key issue to thisgroup. Destination skiers generally favouremployee housing as it will probably resultin better service. Local skiers opposeemployee housing over environmental con-cerns and generally negative attitudestoward resort expansion. Effectively manag-ing resort expansion and operations in sucha contradictory policy environment isextremely difficult.

Management implications

The time and litigation costs of gainingapproval for ski resort expansions hasincreased exponentially over the past twodecades with various stakeholder groups voic-ing concerns and opposing different develop-ments. Understanding these stakeholdergroups is a necessary component of sustain-able tourism development and managementin these mountain resort communities (Gilland Williams, 1994). Thus, the key manage-ment implications of this chapter are theimportance of:

1. Understanding the diversity of the resortenvironment, including not only guests, butalso employees and local residents. It is impor-

tant to go beyond survey research, to employqualitative methodologies such as focusgroups and observational research to under-stand the evolving motivations, concerns andvalues of the different stakeholder groups;2. Developing expansion plans, managementstrategies, and marketing campaigns with aclear understanding of the needs and con-cerns of the different stakeholder groups;and3. Monitoring the satisfaction of each groupwith the resorts’ actions and programmes.

Research directions

Stakeholder research in tourism has focusedheavily on quantitative surveys of host com-munity residents. Similarly, the consumersegmentation research in tourism has alsofocused almost exclusively on quantitativesurvey methodologies. It would be extraor-dinarily difficult to design a survey thatwould capture the population variancedescribed in this paper. Clearly, there is arole for both quantitative and qualitativeresearch programmes. As such, the follow-ing research conclusions and directions areproposed.

1. Long-term observational studies are a valu-able tool to identify and describe the diversityof stakeholder groups in tourism communities.2. It is important to understand and recog-nize at least three populations that character-ize the leisure/tourism setting – theconsumers, resort employees and the hostcommunity residents. It is only through theinteractions of these three groups that we canbegin to understand the true diversity oftourism and leisure settings.3. Understanding the various demonstrationeffects – the relationships between the variousgroups in the resort setting – is important tounderstanding the dynamics of sustainabletourism development. The attitudes of thevarious groups toward each other and towardthe resort corporations are critical.

262 R.R. Perdue


References

Ahn, B.Y., Lee, B.K. and Shafer, C.S. (2002) Operationalizing sustainability in regional tourism planning:an application of the limits of acceptable change framework. Tourism Management 23, 1–15.

Anderek, K. and Vogt, C. (2000) The relationship between residents’ attitudes toward tourism and tourismdevelopment options. Journal of Travel Research 39, 27–36.

Berry, S. and Ladkin, A. (1997) Sustainable tourism: a regional perspective. Tourism Management 18,433–440.

Bowen, D., Schneider, B. and Kim, S. (2000) Shaping service cultures through strategic human resourcemanagement. In: Swartz, T. and Iacobucci, D. (eds) Handbook of Services Marketing and Management.Sage Publishing, Thousand Oaks, California, pp. 439–454.

Brown, J. and Dev, C. (2000) Improving productivity in a service business: evidence from the hotel indus-try. Journal of Service Research 2, 339–354.

Driver, C. and Johnston, R. (2001) Understanding service customers: the value of hard and soft attributes.Journal of Service Research 4, 130–139.

Fay, A. (2000) A History of Skiing in Colorado. Western Reflections Inc., Ouray, Colorado.Fick, G. and Brent Ritchie, J.R. (1991) Measuring service quality in the travel and tourism industry. Journal

of Travel Research 30, 2–9.Freeman, R.E. (1994) Strategic Management: a Stakeholder Perspective. Prentice Hall, Englewood Cliffs, New

Jersey.Friedman, A.L. and Miles, S. (2002) Developing stakeholder theory. Journal of Management Studies 39, 1–21.Garrod, B. and Fyall, A. (1998) Beyond the rhetoric of sustainable tourism. Tourism Management 19,

199–212.Gill, A. and Williams, P. (1994) Managing growth in mountain tourism communities. Tourism Management

15, 212–220.Hunter, C. (1997) Sustainable tourism as an adaptive paradigm. Annals of Tourism Research 24, 850–867.Jackson, G. and Morpeth, N. (1999) Local Agenda 21 and community participation in tourism policy and

planning: future or fallacy. Current Issues in Tourism 2, 1–38.Jones, T.M. and Wicks, A.C. (1999) Convergent stakeholder theory. Academy of Management Review 24,

206–221.Kotler, P., Bowen, J. and Makens, J. (1999) Marketing for Hospitality and Tourism, 2nd edn. Prentice Hall,

Upper Saddle River, New Jersey.Ledgerwood, C., Crotts, J. and Everett, A. (1998) Antecedents of employee burnout in the hotel industry.

Progress in Tourism and Hospitality Research 4, 31–44.Lipsher, S. (2002) Ski resort unveils expansion. Denver Post, 11 September 2002McCool, S.F., Moisey, R.N. and Nickerson, N.P. (2001) What should tourism sustain? The disconnect with

industry perceptions of useful indicators. Journal of Travel Research 40, 124–131.Miller, G. (2001) The development of indicators for sustainable tourism: results of a Delphi survey of

tourism researchers. Tourism Management 22, 351–362.Noe, F. (1999) Tourist Service Satisfaction. Advances in Tourism Applications, vol 5. Sagamore Publishing,

Champaign, Illinois.Oliver, R. (1997) Satisfaction: a Behavioral Perspective on the Consumer. McGraw Hill, New York.Perdue, R. (2000) Service quality in resort settings: trends in the application of information technology.

In: Gartner, W. and Lime, D. (eds) Trends in Outdoor Recreation, Leisure and Tourism. CAB International,Wallingford, pp. 357–364.

Perdue, R., Long, P. and Kang, Y. (1999) Boomtown tourism and resident quality of life: the marketing ofgaming to host community residents. Journal of Business Research 44, 165–178.

Pigram, J. (1990) Sustainable tourism: policy considerations. Journal of Tourism Studies 1, 209.Roos, I. (2002) Methods of investigating critical incidents: a comparative review. Journal of Service Research 4,

193–204.Ryan, C. (2002) Equity management, power sharing and sustainability: issues of the ‘new’ tourism. Tourism

Management 23, 17–26.



Sautter, E.T. and Leisen B. (1999) Managing stakeholders: a tourism planning model. Annals of TourismResearch 26, 312–328.

Shane, A.M. and Graedel, T.E. (2000) Urban environmental sustainability metrics: a provisional set. Journalof Environmental Planning and Management 43, 643–663.

Smith, V.L. (2001) Sustainability. In: Smith, V.L. and Brent, M. (eds) Hosts and Guests Revisited: TourismIssues of the 21st Century. Cognizant Communications, New York, pp. 187–200.

Snepenger, D., Johnson, J. and Rasker, R. (1995) Travel stimulated entrepreneurial migration. Journal ofTravel Research 34, 40–44.

Spiselman, A. (1995) Visionary Ritz-Carlton takes TQM to new level. Hotels October, 45–46.WBCSD (World Business Council for Sustainable Development). The Business Case for Sustainable

Development: Making a Difference Towards the Earth Summit 2002 and Beyond. CorporateEnvironmental Strategy 9, 226–235.

WCED (World Commission on Environment and Development). Our Common Future. Oxford UniversityPress, London.

Yuksel, A. and Yuksel, F. (2001) Measurement and management issues in customer satisfaction research:review, critique, and research agenda. Journal of Travel and Tourism Marketing 10, 47–80.

Zeithaml, V.A. and Bitner, M.J. (2000) Services Marketing, 2nd edn. McGraw Hill, New York.

264 R.R. Perdue


Chapter twenty-threeCultural Determinants of Tourist Intention to Return

Jeffery M. CaneenBrigham Young University – Hawaii, Laie, HI 96762, USA

Abstract

Repeat visitation is one aspect of economically sustainable tourism. While there is a significant body ofresearch relative to travel motivation and destination choice, it has largely focused on the first-time leisurevisitor. This chapter considers the criteria by which tourists decide to return to a destination. It also analy-ses the effect of culture and nationality on the intention to return. A comparison of US, Japanese andChinese tourists in Hawaii found that there are significant differences in their decision making criteria rel-ative to repeat visitation. Japanese tourists indicate a high intent to return to destinations that are fun andrelaxing, while US visitors indicate intent to return based on a desire to learn more about the culture andthe people. Chinese visitors showed greater similarity to US than to Japanese tourists. The study also foundthat reported intent does not always correspond to actual behaviour. Japanese visitors report higher intentthan US visitors to return to Hawaii, while US visitors actually return at a much higher rate.

Introduction

Tourism destination residents often perceivetheir guests as embodiments of culturalstereotypes. In Hawaii, for example, the pre-dominant travel segments are composed ofmainland US ‘westbound’ travellers andJapanese ‘eastbound’ travellers (Hawaii,2000). For locals, the image of the big-spend-ing, non-tipping, low-impact, picture-takingJapanese group tourist stands in contrast tothe laid back, bargain-hunting, adventure-seeking, sometimes obnoxious, mainland‘haole’, Foreign Independent Tour (F.I.T.)tourist. Research has largely confirmed thestereotypes (Sheldon and Fox, 1988; Morris,

1990; Keown, 1989). This study was under-taken to gain insight into the significance ofthese cultural differences on a tourist’s deci-sion to return to a destination.

The great majority of research relative totravel motivation and destination choice hasfocused on the first-time leisure visitor.Destination novelty is both an inducementto travel (Lee and Crompton, 1992) and abarrier. Illusion and anticipation, it hasbeen said, is the essence of tourism (Fakeyeand Crompton, 1991). On the other hand,first-time visitors often feel considerableanxiety about their choices. Researchershave recognized that a vacation represents ahigh-risk expenditure for travellers due to



the unknowns of travelling to a new destina-tion (Mansfield, 1992). Repeat visitationwould seem to alter the risk/reward calcula-tion of vacation travel. How does this affectmotivation? On what basis do travellersdecide to return?

Repeat visitation

Cohen (1972) and Schmoll (1977) bothfound that destination experience is only aminor factor in the travel purchase decision.Milman’s widely referenced, multi-step travelbuying process credits destination experienceas a filter through which future vacationchoices are either broadened or narrowed. Afavourable experience might lead to thechoice of a similar destination for future vaca-tions, but will not necessarily result in repeatvisitation (Milman, 1993). Experience doesappear to make the traveller more attentiveto the advertising and promotional efforts ofthe destination (Sonmez and Graefe, 1995).The literature indicates that the illusion of ayet-to-be-visited destination rarely coincideswith the reality of experience. Tourists reportsignificant shifts in every aspect of destinationimage after their visit (Fakeye and Crompton,1991). They do, however, return. Successfulmass tourism destinations depend on repeatvisitors. Fully 58% of tourists vacationing inHawaii, for example, are repeat visitors(Hawaii, 2000).

Why do they return? The research identi-fies the desire to reduce risk and the assur-ance of dealing with nice people in acomfortable, ‘at home’ environment. Alsomentioned is the desire to see sights missedon previous visits and to introduce friendsand family to the destination (Gitelson,1984). Oppermann (2000) noted that visitfrequency itself is a good predictor of futurevisitation. The more frequently the destina-tion was visited in the past, the more likely itwill be visited again.

Repeat visitation can be seen as a type ofproduct or brand loyalty. A common researchfinding on this topic is that satisfaction with aproduct is an unreliable predictor of repeatpurchase (Reichfield, 1996; Stewart, 1997).Tourism destinations and individual firms com-

monly seek to assess traveller satisfaction, pre-sumably to determine the prospect of a returnvisit. Loyalty, however, does not appear to becreated by so simple a formula (Oliver, 1997).

Obstacles to the loyalty-building processinclude variety seeking and multi-brand loy-alty (Oliver, 1999). As variety is a major travelmotivation, this would seem to be a ratherformidable constraint to repeat visitation.

Cultural variations of tourist behaviour

There are observable differences in the travelbehaviour of tourists based on nationality(Yuan and McDonald, 1990). Most of the vari-ation in behaviour can be explained in termsof the cultural values of the collective ‘good’versus the individual ‘good’ (Triandis, 1991).This study is specifically concerned with com-paring the behaviour of North American,Japanese and Chinese tourists. A few studieshave focused on the motivation and behav-iour of US tourists compared to Japanesetourists. Sheldon and Fox (1988) describedJapanese tourists as more reluctant to try newcuisine. They appeared to place greaterimportance on quality service and ‘high-end’dining than American tourists. Japanesetourists appear to make destination choicesbased on their perceptions and expectationsof natural beauty, good beaches, historicspots, modern urban culture, good shoppingand personal safety (Morris, 1990).

In an often-cited study, Keown (1989)identified unique travel behaviour that hasbecome the stereotype of the Japanesetourist. He identified the emphasis on shop-ping, the unique cultural obligation ofomiyage, the ubiquity of picture taking, andthe short-duration, big-spending vacation thattypifies the Japanese tourist and makeshim/her a highly sought after customer bydestinations around the world. Yamamotoand Gill (1999) explained this behaviour bynoting that the Japanese cultural values oftogetherness and social interaction play alarge part in their vacation choices. In con-trast, they found that North Americanschoose a travel destination based primarily onthe desire to get away, relax and experience anew culture.

266 J.M. Caneen


In spite of the fact that increasing free-dom and affluence in China has the poten-tial to produce millions of new leisuretravellers, there has been a scarcity ofresearch regarding the travel motivations oftourists from China, Hong Kong and Taiwan.There is also a lack of research on the ques-tion of variations in destination loyalty by cul-tural background.

Current Study

This study had multiple objectives: (i) togather information regarding destinationcharacteristics that are likely to result inrepeat visitation; (ii) to determine if thosecharacteristics differ based on the culturalbackground of the tourist; (iii) to comparetourist return criteria with tourist impressionsof a popular destination (Hawaii) and a pop-ular attraction in that destination (thePolynesian Cultural Center, PCC); and (iv) tocompare the survey data to actual repeat visi-tation statistics for Hawaii and the PCC.

Identifying and influencing prospectivevisitors is the primary purpose of destinationpromotional organizations. It is also the mar-keting objective of individual firms. Potentialrepeat visitors are comparatively easy andinexpensive to identify, and they are morereceptive to the marketing message.Destinations and attractions would do well tounderstand the criteria by which travellersmake the choice to return. This studyattempts to identify those criteria.

Methodology

A structured-interview survey instrument wasdesigned and administered to 428 randomlyselected visitors to the PCC. Equal numbersof survey attempts were made each eveningfor 1 week during the spring tourism seasonof 2001. The interview included questionsdesigned to segment the sample by countryof origin, age, gender and visit frequency.Respondents were asked to assess their levelof agreement (on a five-point scale) withstatements regarding criteria for returning toa destination as well as their perceptions of

both Hawaii and the PCC in relation to thesame criteria. In each case, a same-native-lan-guage interviewer administered the survey.

The basic purpose of each statement wasto determine the relative effect of each of thefollowing on a visitor’s intention to return toa leisure travel destination:

1. Attractive shopping opportunities.2. Opportunity to interact with the residentpopulation and experience a different cul-ture.3. Opportunity to spend time with familymembers.4. Desire to show family and friends a desti-nation previously visited.5. Desire to do or see things that were missedon a previous visit. A perception of noveltyand change – not just the ‘same old thing’.6. A welcome and comforting feeling ofbeing ‘at home’ at the destination.7. The perceived value of the destination.8. Opportunity for active recreational pur-suits.9. Opportunity for thrilling or risky activities.10. Opportunity for relaxing activities.11. The perceived level of service.12. Visitor density or level of ‘crowding’.13. Perceived physical beauty of the destina-tion.

The total sample size was 428. Residents ofthe USA (191 respondents), Japan (148respondents) and China (54 respondentsfrom both the PRC and Taiwan) made up asignificant majority of the sample. Because ofinsignificant representation, responses fromother countries were deleted, resulting in 393usable interviews.

Experience has shown that Japaneserespondents are reluctant to complainbecause of politeness or empathy, and that‘even objective measures may show little varia-tion, or are often clustered about the mean… non-parametric measures such as rankings… might be necessary to discern subtle differ-ences (or differences subtly expressed)’(Iverson, 1997). Accordingly, the surveyresponses were segmented by country of ori-gin and the criteria were subjected to thenon-parametric Kendall’s W test, whereby aweighted rank was assigned to each. TheKendall’s W rankings were tested by means of

Cultural Determinants of Tourist Intention to Return 267


a one-way analysis of variance (ANOVA) todetermine variation (if any) according tocountry of origin. The Bonferroni test wasconducted to determine variation betweenany two of the three countries. The results ofthese tests can be seen in Tables 23.1–23.3.

To compare the relevance of nationality inthe decision to return to that of other demo-graphic variables, similar rankings andANOVA tests were made based on the ageand gender of respondents.

Results

The decision to return

Japanese visitors indicated that their intent toreturn is most strongly influenced by: (i) theopportunity to get away and relax; (ii) theperceived natural beauty of the destination;and (iii) the opportunity to spend time withfamily. US visitors based their intent to returnprimarily on: (i) the opportunity to spend

268 J.M. Caneen

Table 23.1. Destination return criteria by visitor nationality.

Overall Japan USA China ANOVA BonferroniKW KW KW KW Sig. significance (%)

Criteria rank rank rank rank (%) 1–2 1–3 2–3

Shopping 8.6 9.4 8.3 6.8 7.9Locals and culture 10.5 10.1 10.7 10.4 2.9 3.1Family 11.4 10.9 12.0 10.7 0.3 0.5Show 11.5 10.7 11.8 11.8 0.6 0.4Novelty 8.9 9.0 9.0 9.0 68.7At home 10.1 9.1 10.7 10.1 0.0 0.0Value 9.6 9.7 9.5 9.8 69.6Active 7.5 8.9 6.7 6.7 1.5 1.6Thrill 6.2 7.2 5.5 7.0 3.3Relax 11.4 11.5 11.4 10.7 76.1Service 11.2 10.6 11.7 10.9 9.3Not crowded 8.8 9.0 9.1 7.4 6.4Beauty 10.7 11.0 10.5 10.9 86.4

KW, Kendall’s W.

Table 23.2. Perceptions of Hawaii by visitor nationality.



Shopping 7.9 8.2 7.6 7.6 79.2Locals and culture 7.5 7.0 7.4 8.8 2.4 3.7Family 10.0 10.7 9.5 9.9 45.1Show 10.4 10.3 10.4 10.6 21.3Novelty 8.7 8.9 8.9 6.2 0.0 0.0 0.1At home 8.0 6.6 9.0 8.1 0.0 0.0Value 6.5 7.1 6.0 7.9 46.0Active 8.1 7.5 8.5 8.1 0.2 0.2Thrill 7.5 7.8 7.6 7.1 25.3Relax 9.4 9.5 9.2 9.1 29.0Service 8.5 8.8 8.2 8.2 53.9Not crowded 5.8 6.1 5.9 6.1 93.4Beauty 11.3 11.0 11.3 11.7 7.4

KW, Kendall’s W.


time with family; (ii) a desire to show friendsand family a destination previously visited;and (iii) the perceived level of service qualityat the destination. Chinese visitors ranked thecriteria by: (i) a desire to show the destina-tion to others; (ii) the perceived service level;and (iii) the natural beauty of the area.

Japanese travellers placed significantlygreater importance on the availability ofactive and ‘thrilling’ activities than did US vis-itors. US tourists, on the other hand, placedsignificantly more importance on interactingwith the local culture, spending time withfamily, showing the destination to others andfeeling ‘at home’ at the destination. Thereturn criteria of Chinese visitors were, some-what surprisingly, more similar to those of USvisitors than to those of Japanese visitors.

Perceptions of Hawaii

Significant differences in the perceptions ofHawaii were discovered as follows:

• Compared to Japanese visitors, US visitorsperceived a greater opportunity to interactwith local people and experience a uniqueculture.

• Chinese visitors experienced Hawaii asbeing more novel and unique than dideither Japanese or American visitors.

• US visitors felt more ‘at home’ thanJapanese visitors.

• Japanese visitors perceived less opportu-nity for ‘thrilling’ activities than did USvisitors.

Graphic comparisons of highest-level returncriteria versus perceptions of Hawaii for eachnationality are shown in Figs 23.1–23.3.

Perceptions of the PCC

Significant differences in the perceptions ofthe PCC were discovered as follows:

• US visitors perceived the shopping oppor-tunities to be significantly more attractivethan did either the Japanese or Chinesevisitors.

• Both US and Chinese visitors reportedfeeling more ‘at home’ than did theJapanese visitors.

• Japanese visitors perceived less opportu-nity for active pursuits than did theAmericans or the Chinese.

• US visitors considered the service at thePCC to be of a higher level than did theJapanese visitors.

• US visitors also rated the physical beautyof the PCC significantly higher than didthe Japanese.

Graphic comparisons of return criteria versusperceptions of the PCC for each nationalityare shown in Figs 23.4–23.6.


Table 23.3. Perceptions of PCC by visitor nationality.



Shopping 5.6 5.5 5.9 4.4 0.1 0.1 4.3Locals and culture 10.7 10.7 10.6 11.0 4.6Family 10.3 10.9 10.0 10.1 95.6Show 10.2 10.6 10.2 9.5 52.0Novelty 7.5 8.4 7.1 6.5 17.6At home 7.9 6.4 8.7 8.6 0.0 0.0 0.0Value 7.4 8.3 6.7 8.4 17.6Active 6.4 5.1 6.7 8.8 0.0 0.0 0.0Thrill 5.7 6.3 5.6 5.6 57.5Relax 8.3 8.3 8.1 9.1 13.0Service 10.6 10.4 10.7 10.0 0.9 0.8Not crowded 6.1 6.7 5.8 6.1 69.8Beauty 11.0 10.8 11.1 10.6 1.5 1.2

KW, Kendall’s W.


Nationality compared with otherdemographic factors

The results of the survey show that there isgreater variation in the ranking of return cri-teria based on nationality than on either the

age or gender of survey respondents. Thesame is true regarding perceptions of thePCC. Perceptions of Hawaii, however, variedmore greatly by age than by either gender ornationality (Table 23.4).

270 J.M. Caneen

6.0

7.0

8.0

9.0

10.0

11.0

12.0

6.0 7.0 8.0 9.0 10.0 11.0 12.0

Importance

Per

cept

ion

Relax

BeautyFamily

Show

Fig. 23.1. Japanese perceptions of Hawaii relative to their highest-ranking return criteria.

6.0

7.0

8.0

9.0

10.0

11.0

12.0

6.0 7.0 8.0 9.0 10.0 11.0 12.0

Importance

Per

cept

ion

Relax

Service

Family

Show

Fig. 23.2. US perceptions of Hawaii relative to their highest-ranking return criteria.

6.0

7.0

8.0

9.0

10.0

11.0

12.0

6.0 7.0 8.0 9.0 10.0 11.0 12.0

Importance

Per

cept

ion

Relax

Service

Beauty

Show

Fig. 23.3. Chinese perceptions of Hawaii relative to their highest-ranking return criteria.


Conclusions

As tourism destinations and commercialattractions attempt to build repeat visitation,their marketing efforts should consider thecultural/national differences of visitors.

Nationality, more than age or gender, signifi-cantly influences the decision to return to atravel destination.

Pacific Rim destinations should be awareof the different return decision criteria usedby Asian and American tourists. Japanese


6.0

7.0

8.0

9.0

10.0

11.0

12.0

6.0 7.0 8.0 9.0 10.0 11.0 12.0

Importance

Per

cept

ion

Relax

BeautyFamily

Show

Fig. 23.4. Japanese perceptions of the PCC relative to their highest-ranking return criteria.

6.0

7.0

8.0

9.0

10.0

11.0

12.0

6.0 7.0 8.0 9.0 10.0 11.0 12.0

Importance

Per

cept

ion

Relax

Service

FamilyShow

Fig. 23.5. US perceptions of the PCC relative to their highest-ranking return criteria.

6.0

7.0

8.0

9.0

10.0

11.0

12.0

6.0 7.0 8.0 9.0 10.0 11.0 12.0

Importance

Per

cept

ion

RelaxService

Beauty

Show

Fig. 23.6. Chinese perceptions of the PCC relative to their highest-ranking return criteria.


tourists in particular seem to be more inter-ested in returning to a destination that theyfound to be relaxing and fun, whileAmerican, and to a lesser extent, Chinesetravellers are more interested in experiencingand learning more about the destination, andin spending time with loved ones.

Finally, as applied to Hawaii as a destina-tion and the PCC as a tourist attraction, thisresearch shows that all three groups reportequal satisfaction with their experience over-all, but does satisfaction create intent toreturn? Here the data present a curious pic-ture. Ninety-six per cent of Japanese touristssay they intend to return to Hawaii. Thereported intent to return of American andChinese tourists is similarly high, but whilethe Japanese report near unanimous intentto return to both Hawaii and the PCC, a sta-tistical analysis shows no significant correla-tion between their intent and their

perceptions of the fulfilment of their high-est-level return criteria. There is also littlecorrelation between the perception andintent of Chinese travellers. Americans, onthe other hand, report a lower intent toreturn but a highly significant correlationbetween intent and their perceptions of thefulfilment of their highest-level return crite-ria (Table 23.5).

Statistics provided by the HawaiianTourism Authority show that westbound trav-ellers are in fact much more likely to visitagain than are eastbound travellers (Hawaii,2000). This research likewise found that 47%of US respondents were repeat visitors,whereas only 36% of Japanese and 13% ofChinese respondents had been to Hawaiibefore. Repeat visitation to the PCC shows anearly identical pattern.

So while eastbound Japanese and Chinesevisitors report a higher intent to return, west-

272 J.M. Caneen

Table 24.4. ANOVA significance (%) of variation in return criteria by age and gender.

Variation by age Variation by gender

ReturnPerceptions

ReturnPerceptions

criteria Hawaii PCC criteria Hawaii PCC

Shopping 13.6 0.5 32.6 1.0 0.5 59.1Locals and culture 95.3 0.0 76.8 17.3 56.8 23.3Family 57.5 0.0 11.2 56.7 18.3 34.6Show 79.0 0.0 17.6 28.7 45.8 18.2Novelty 8.0 0.0 8.7 71.0 48.2 27.0At home 51.4 0.0 0.2 22.8 35.2 61.6Value 85.3 21.8 70.3 79.3 29.8 29.5Active 0.0 0.5 52.9 21.9 10.7 34.9Thrill 0.0 0.0 6.9 8.9 19.2 8.4Relax 27.2 0.1 24.9 63.9 38.7 82.1Service 4.9 0.3 16.3 34.5 39.3 83.3Not crowded 16.3 8.0 36.8 42.5 90.1 84.6Beauty 7.9 0.0 4.1 15.4 61.6 10.1

Table 23.5. Significance of the correlation between perceptions of highest rank return criteria and thereported intent to return.

Significance of correlation (%)

Return intent (%) Return actual (%) 1st rank 2nd rank 3rd rank

Hawaii PCC Hawaii PCC Hawaii PCC Hawaii PCC Hawaii PCC

Japan 96 96 36 35 0.604 0.710 0.313 0.138 0.412 0.154USA 91 76 47 49 0.022 0.000 0.010 0.000 0.114 0.002China 98 89 13 4 0.021 0.583 0.109 0.817 0.322 0.162


bound US visitors actually return at a muchhigher rate. The discrepancy betweenreported intent and observed behaviourcould be attributed to the aforementionedreluctance of Japanese to criticize or disap-point – making the survey results somewhatunreliable. Or it might be that their intent isindeed higher but their ability to return islower. Future research on this questionshould incorporate the use of a broadermeasurement scale – a ten-point rather thana five-point scale would better accommodatethe subtleties of opinion encountered in thisresearch. It would also be interesting to sur-

vey Japanese, Chinese and American touristson their actual vacation destination deci-sions over a period of years.

This research has additional limitationsthat should be considered in the interpreta-tion of its findings. The study sample is smalland is limited to a relatively specific group oftravellers who visited the PCC. It is likely thatthose who choose to visit the PCC are notentirely representative of tourists visitingHawaii. It is even more unlikely that the sam-ple results can be generalized to the largerpopulation of American, Japanese andChinese tourists.


References

Cohen, E. (1972) Toward a sociology of international tourism. Social Research 39, 164–182.Fakeye, P. and Crompton, J. (1991) Image differences between prospective, first time and repeat visitors to

the Lower Rio Grande Valley. Journal of Travel Research 30, 10–19.Gitelson, R.J. (1984) Insight into the repeat vacation phenomenon. Annals of Tourism Research 11, 199–217.Hawaii (2000) State of Hawaii – Development of Business, Economic Development and Tourism, 1999 Annual Visitor

Research Report, December 2000.Iverson, T. (1997) Japanese visitors to Guam: lessons from experience. Journal of Travel and Tourism

Marketing 6, 41–54.Keown, C.F. (1989) A model of tourist’s propensity to buy: the case of Japanese visitors to Hawaii. Journal of

Travel Research 28, 31–34.Lee, T. and Crompton, J. (1992) Measuring novelty seeking in tourism. Annals of Tourism Research 19,

732–751.Mansfield, Y. (1992) From motivation to actual travel. Annals of Tourism Research 19, 399–419.Morris, S. (1990) The Japanese Overseas Travel Market in the 1990’s. Economist Intelligence Unit, London.Oliver, R.L. (1997) Satisfaction: A Behavioral Perspective on the Consumer. McGraw-Hill, New York.Oliver, R.L. (1999) Whence consumer loyalty? Journal of Marketing 63, 33–44.Oppermann, O. (2000) Tourism destination loyalty. Journal of Travel Research 39, 78–84.Reichheld, F.F. (1996) Does satisfaction moderate the association between alternative attractiveness and

exit intention in a marketing channel? Journal of the Academy of Marketing Science 22, 364–371.Schmoll, G.A. (1977) Tourism Promotion. Tourism International Press, London.Sheldon, P.J. and Fox, M. (1988) The role of food service in vacation choice and experience: a cross-

cultural analysis. Journal of Travel Research 27, 9–15.Sonmez, S.F. and Graefe, A.R. (1995) International vacation decisions and the threat of terrorism.

Proceedings of the Annual Conference of Travel and Tourism Research Association 26, 236–245.Stewart, T.A. (1997) A satisfied customer isn’t enough. Fortune 136, 112–113.Triandis, H.C. (1994) Culture and Social Behavior. McGraw-Hill, New York.Yamamoto, D. and Gill, A. (1999) Emerging trends in Japanese package tourism. Journal of Travel Research

38, 134–143.Yuan, S. and McDonald, C. (1991) Motivational determinates of international pleasure time. Journal of

Travel Research 29, 42–43.


Chapter twenty-fourTowards the Conceptualization of Tourism DestinationLoyalty

Outi Niininen and Michael RileySchool of Management, University of Surrey, Guildford, Surrey GU2 7XH, UK

Abstract

Consumer loyalty is an attractive field of research to managers in the retail industry and recently therehave been studies into the tourists’ propensity to spend consecutive vacations in one specific destination.This chapter proposes a conceptualization for destination loyalty by following the route set out by Jacobyand Chestnut (Brand Loyalty Measurement and Management, John Wiley & Sons, Chichester, 1978). The pro-posed destination loyalty concept also incorporates the psychological need for consistency (and variety) ineveryday life by acknowledging the motivational power of the optimum stimulation level (OSL).

Introduction

Tourism destination loyalty has been the sub-ject of numerous studies; Gyte and Phelps(1989) as well as Ryan (1995) studied touristsreturning to Mallorca, and Oppermann(2000) investigated repeat vacations betweenNew Zealand and Australia. The existingresearch shows that destination loyal behav-iour does exist, yet many of the findings arenon-comparable due to the differences inconceptualization of the loyalty construct(Oppermann, 2000).

Consumer loyalty has long been viewed asan ‘enigma’ to researchers (Snepenger,1987). Traditionally business managers haveassumed that the popularity of a brand is sim-ply due to effective distribution and superior

product features. However, this does notexplain the phenomenon that some individu-als are prepared to endure discomfort intheir purchasing process rather than switchbetween brands nor that some customers arenot prepared to change their purchasing pat-terns even when a competing product offersgreater utility (Snepenger, 1987). In otherwords, some consumers have a higher psycho-logical propensity to become a loyal buyer ofa brand (Reichheld and Teal, 1996;Srinivasan, 1996). Furthermore, destinationloyalty as a phenomenon has been identifiedby Mazurski (1989), Gyte and Phelps (1989),Ryan (1995), Baloglu and Erickson (1998),Oppermann (1998, 1999, 2000) as well asDarnell and Johnson (2001). Yet, the incon-sistency in the conceptualization of loyalty has



limited the comparison of these various stud-ies (Dekimpe et al., 1997; Oppermann, 2000).

Tourism is an interesting field for loyaltyresearch, since the product is intangible, rela-tively expensive, purchased infrequently andcan be extremely important to the consumer.This type of product may attract very strongloyalty as well as a clear correlation to demo-graphic variables. Further, once loyalty hasbeen established, it is perceived as fairly con-sistent over time (Bloemer et al., 1999).Conversely, tourism may be an unlikely prod-uct for loyalty due to its variety seeking andplay characteristics: ‘it’s voluntary, outsideordinary life, limited in time and space, sur-rounded by an air of mystery, utterly absorb-ing yet recognized as being somewhatmake-believe, it has something at stake, anoutcome in doubt’ (Godbey and Graefe,1991, p. 214). Furthermore, ‘the novelty driveis an underlying motivation for pleasuretravel’ (Crotts, 1993, p. 7).

The inconsistent conceptualization of des-tination loyalty has been acknowledged(Oppermann, 2000). The description oftourism as a form of play (Godbey andGraefe, 1991) draws attention to the psychol-ogy concepts of exploratory behaviour and vari-ety seeking. Exploratory behaviour is the way bywhich individuals control their perceivedlevel of stimulation from the surroundingenvironment; such a tendency to aspire to amedium degree of stimulation is called theoptimum stimulation level (OSL) (Baumgartnerand Steenkamp, 1994). The OSL concept isappropriate for this study since stimulation-seeking behaviour has been under systematicresearch since the 1950s and the optimalstimulation level is based on personality traits(Venkatesan, 1973; Arnett, 1994, 1996; Rileyet al., 2001; Niininen, 2002). Furthermore,this seeking of OSL also embraces variety, nov-elty or sensation seeking as well as curiosity (typesof exploratory behaviour), all of which arewell suited to the holiday behaviour context,especially when the reward potential ofincreased stimuli is recognized (Fiske andMaddi, 1961; Kish and Donnenwert, 1969;Wahlers and Etzel, 1985).

Variety seeking is intrinsically motivatedbehaviour where a change from routine isrewarded by pleasure; hence, variety seeking

is the primary motivational mechanism main-taining the OSL (Rogers, 1979; Menon andKhan, 1995). Variety seeking in buying behav-iour depends on the product characteristicslike involvement, purchase frequency, hedo-nic characteristics of the brand and the per-ceived differences between brands (van Trijpet al., 1996). Variety seeking in the tourismcontext has been researched by Bello andEtzel (1985), Lee and Crompton (1992) andSnepenger (1987).

The focus here is on loyalty as a humanpsychological and behavioural phenomenon;hence, the emphasis is on identifying theinternal processes and reasons an individualmight have for loyal buying behaviour, not onthe management of customer retention. Thischapter has the following structure: initiallythe various definitions of loyalty are discussedand the consumer loyalty as a phenomenon isrelated to the tourism industry; finally theproposed model for inherent destination loy-alty is presented.

Loyalty Definitions

Brand loyalty can be defined at three specificlevels: (i) cognitive behaviour, where loyalty is‘an internal commitment to purchase and re-purchase a particular brand’ (Evans et al.,1996, p. 261); (ii) an attitudinal phenomenonwhere the consumer has a preference to use aparticular brand; and (iii) a behavioural phe-nomenon where brand loyalty is repeat pur-chases of a particular brand. However, it isimportant to recognize that any habitual pur-chases of a brand must have initially resultedfrom cognitive processes and therefore theexistence of ‘pure’ behavioural loyalty can bequestioned (Evans et al., 1996). In fact, theword ‘loyalty’ is widely (mis)used to describemany aspects of repetitive buying behaviour(Uncles and Laurent, 1997). Consequently,the term has suffered from inflation and isnow deeply embedded in marketing jargon torepresent any repetitive buying pattern. Thismakes the comparison of past research espe-cially complicated.

This paper proposes a conceptual defini-tion of brand loyalty as ‘the biased (i.e. non-random) behavioural response (i.e.

276 O. Niininen and M. Riley


purchase), expressed over time, by some deci-sion making unit, with respect to one or morealternative brands out of a set of such brands,and it is a function of psychological (decisionmaking, evaluative) process’ (Jacoby andChestnut, 1978, p. 80). This definition of loy-alty recognizes two dimensions: psychologicalattachment and behavioural consistency (Day,1969; Jacoby and Kyner, 1973; Howard et al.,1988; Backman and Crompton, 1991). Therelationship between these two constructs isdemonstrated in Fig. 24.1.

The benefit of applying this conceptualdefinition is that it identifies the latent loyals,i.e. those with genuine preference for abrand but no consistent purchase history, aswell as spurious loyals, those with a high behav-ioural consistency combined with a low psy-chological attachment towards the brand.Latent loyalty could be associated with expen-sive brands that attract admiration from allsegments of society, regardless of the fewopportunities to purchase such items (e.g. avacation on a private island in theCaribbean). By contrast, an example of spuri-ous loyalty would be individuals who have lim-ited access to the brands they prefer andconsequently have to ‘settle with’ alternativeproducts (e.g. no time for long-haul holiday).In most cases the high and low loyals are theeasiest to classify since they are at the oppo-site ends of the loyalty continuum.

One interpretation of this psychologicalattachment is based on attitudes. Accordingto Dick and Basu (1994), this brand attitudeis divided into cognitive, affective and conativecomponents. The cognitive antecedent is con-cerned with the information the individualholds about a brand and how easily he/shecan recall this data from memory (this

depends on the relative strength of the associ-ation between the data and the brand). Theaffective antecedents explain the feelings theindividual holds about the brand; these emo-tions can actually be a stronger predictor offuture behaviour than a cognitive evaluation,especially if the behaviour becomes a habitrequiring very little conscious deliberation(Dick and Basu, 1994). This could refer tothe ‘friendship’ the loyal customer is likely todevelop with their usual purchase (Driver,1996). The conative component includes eval-uation of switching costs and sunk costs, e.g. ini-tial fees to join the service, the difficulty offinding a quality alternative and the expecta-tions of the brand’s performance. In otherwords, the stronger a person’s attitudetowards a brand the harder he/she is pre-pared to work to overcome any obstacles touse or re-purchase this brand.

There are over 50 operational definitionsfor loyalty, most of which focus on behaviouralloyalty. These measures emphasize the propor-tion of product category expenditures on agiven brand (Jacoby and Chestnut, 1978).Behavioural approaches account for approxi-mately 60% of known loyalty measures.Additionally, the variety of operationalizedmeasures makes it very difficult to compareresults and draw holistic conclusions. Finally,behavioural measures fail to distinguishbetween habitual buying and true loyalty(Olson and Jacoby, 1971; Backman andCrompton, 1991; Uncles and Laurent, 1997).

Simple behavioural measures of loyalty leadto overestimations of the loyal following forany brand because behavioural measures donot discount spuriously loyal buyers.Therefore, to address fully the conceptualiza-tion of brand loyalty, a composite measure

The Conceptualization of Tourism Destination Loyalty 277

Psychological attachment

Behaviouralconsistency

Low loyalty

Spurious loyalty

Latent loyalty

High loyalty

Low

High

Weak Strong

Fig. 24.1. Loyalty as a psychological and behavioural construct. Source: Backman and Crompton (1991, p. 3).


(with both behavioural and attitudinal compo-nent) is required (Backman, 1991). In pastresearch, these composite measures haveincluded evaluations of price, brand commit-ment and information search models.However, the attitude-behaviour model is themost commonly applied to tourism and leisureresearch (Oppermann, 1999). Furthermore,the composite measures have been praised fortheir comprehensive approach to brand loy-alty. A loyal attitude towards a holiday destina-tion, e.g. psychological attachment, can also proveimportant to the management of the negativeimpacts of tourism.

Jacoby and Chestnut (1978) were very crit-ical about operational loyalty measures basedon empirical work only, since these measureswere mainly based on arbitrary criteria.Furthermore, there were no attempts toanalyse the opposite type of behaviour, i.e.dis-loyalty. Moreover, all of these measuresfocused on overt buying behaviour, henceacceptance of the Black Box theories, where theresearchers were only interested in the out-come of complex psychological processes.There was no attempt to evaluate the rela-tionship between established measures andthus discover whether they were measuringthe same phenomenon. The research meth-ods used also lacked validity, reliability andsensitivity (Jacoby and Chestnut, 1978). Inother words, many of the past measures hadno conceptual underpinning and only simpli-fied the phenomenon.

The attitudinal measures were generallybased on scales where the strength of theexpression can be quantified; yet the wordingof these statements is open to bias. Overall,the introduction of the attitudinal measureshas improved the sensitivity of the researchinstrument (Jacoby and Chestnut, 1978) andthere are also reports of improved test–re-testresults (Olson and Jacoby, 1971). In conclu-sion, the composite loyalty measures withboth behavioural and psychological elementsallow for the greatest accuracy of the findings.Through this approach, even the spuriousloyal can be identified for analysis (Backmanand Crompton, 1991), and the social pressureto conform and give politically correct answerscan be eliminated by analysing establishedbehavioural patterns.

Loyalty in a Tourism Context

The key reasons for an individual to indulgein repeat tourism are based on the opportu-nity to reduce the perceived risk of a holiday,to meet like-minded people, to build an emo-tional attachment towards the destination(i.e. true loyalty) and an opportunity to fur-ther explore the destination or show it tofriends. Furthermore, tourists seeking relax-ation from their holiday are most likely toreturn to the same destination (Gitelson andCrompton, 1984; Pyo et al., 1998) and themore familiar tourists are with their destina-tion, i.e. habitual tourists or extensive sec-ondary information of the location, the lessplanning time (and effort) is required (Cookand McCleary, 1983; Etzel and Wahlers,1985). The destination image is formed dur-ing the tourist’s first visit to the destinationand this image is only re-emphasized duringfurther visits to the same location(Oppermann, 1997). Therefore, first-time vis-itors also have a tendency to experience moreof the destination, visit more attractions andgenerally seek variety and culture from theirholiday experience, whereas repeat visitorsaim for a relaxing holiday (Pyo et al., 1998).

The natural tendency of not returning tothe same destination has also been presentedin tourism literature with several references towanderlust (Ross, 1994) and tourism travel career(Hitchcock et al., 1994). The travel career con-cept is an attempt to apply Maslow’s hierachyof needs to destination selection situation.Furthermore, in a separate study, Baloglu andErickson (1998) argued that ‘the tendency fora traveller to switch from one destination toanother is greater than the tendency to revisitthe same destination’ (as cited in Pyo et al.,1998, p. 182).

Furthermore, novelty seeking in tourism hasbeen associated with a lower propensity toreturn to a holiday destination (Bello andEtzel, 1985). The need for novelty is per-ceived to be a fundamental element of holi-day; Crompton (1979: as cited in Bello andEtzel, 1985) argues that repeat visits to a desti-nation are only sought ‘when the tourists ismotivated by specific socio-psychologicalmotives such as kinship or social interaction’or anxiety avoiding (as cited in Bello and



Etzel, 1985, p. 21). In other words, noveltyseeking is seen as the opposite of loyal buyingbehaviour in destination selection. Therefore,research into novelty seeking behaviour canadd to the understanding of destination loy-alty. Given that novelty seeking behaviour isan outcome of maintaining optimal individ-ual levels of stimulation, this general propen-sity to moderate stimulation levels can beused to study novelty (and loyalty).

The main reasons that discourage loyaltourism behaviour are the general trends in themarketplace, e.g. increased opportunities toswitch between destinations as ‘new’ vacationlocations are emerging each year. With suchwide selection of substitute holiday destina-tions, the perceived switching costs are lowerand there is a lower perceived risk involvedwith exploring a new holiday destination (Raj,1985; Rivers et al., 1991; Driver, 1996).

At the same time, tourists are benefitingfrom developments in the travel industry.Increased travel experience (whether forwork or leisure) will have an impact ondomestic tourism and short breaks.Furthermore, enhanced travel conveniencewill open up long-distance travel even forthose who were previously unwilling toendure the stress of such journeys. The pro-motion of tourism has also changed.Increased advertising encourages productswitching and the introduction of real timefilm footage of the destination over theInternet or Disney-style destination videosattempts to make the tourism product moretangible, i.e. makes switching between desti-nations easier. To conclude the above, thelevel of perceived risk associated with vaca-tions will decrease for some tourists; however,for many others the main holiday of the yearwill remain a high-involvement purchase.

There have also been significant changeson the supply side of tourism. Tourism desti-nations are now improving their productimage; many destinations are systematicallyintroducing branding to their advertising.Furthermore, destination managers are alsogetting better at segmenting their target mar-ket and positioning different types of destina-tion to ensure a better match between thedesires of the tourists and the destinationfacilities and services. Finally, many destina-

tions are introducing (or implementing)strict product quality improvement throughlicensing and planning laws and so accommo-dating the quality conscious loyal consumers(East et al., 1995).

Having less time to shop around has beenproven to increase loyalty in grocery shopping(Denison and Knox, 1992), and today’s busylifestyle, combined with the experienced levelof arousal, has also been reported to influ-ence the type of holiday chosen. Such a trendmay encourage return visits to destinationsthat have ‘proven their qualities’ in the past.

Higher perceived risk involved with the ser-vice products (and especially tourism) can leadto stronger loyalty when loyalty is seen as arisk-reduction strategy for the service user(Snyder, 1991; Rowley and Dawes, 2000).Therefore, the unique nature of the tourismproduct (high involvement, relatively largeannual investment and high perceived risk)encourages repeat holidays for the majority oftourists (Gitelson and Crompton, 1984; Dickand Basu, 1994; Pritchard and Howard, 1997).Finally the demographic trend of an agingpopulation also encourages loyalty sinceolder consumers choose goods from a nar-rower band (Uncles and Ehrenberg, 1990).

To conclude, loyalty research in tourism isattractive since the product under focus ispurchased regularly but infrequently, andconstitutes a large part of the consumers’household budget but is not a consumerdurable. This can lead to a ‘fiercely loyal fol-lowing even though the customer’s loyalty iscashed in less often’ (Exter, 1986, p. 33).Therefore, the type of loyalty being sought isunrelated to daily life and carries a long life-span. Moreover, for a high involvement prod-uct (such as tourism) the actual productconsumption (the holiday) offers the cus-tomer more satisfaction than any loyaltyscheme incentive ever could (Dowling andUncles, 1997), which undermines the value ofsuch schemes for tourism business.

However, it must be reiterated that anyresearch into tourism destination loyalty isproblematic since brand loyalty as behaviourcannot be generalized across product cate-gories (Charlton, 1973) and the relationshipbetween customer evaluations and loyalbehaviour between different industries is ‘nei-



ther linear nor simple’ (Cronin and Taylor,1992; Parasuraman et al., 1994; Jones andSasser, 1995). To conclude, this study has aunique focus and it should be treated asexploratory research.

In the model proposed here, the burdenof defining a ‘destination’ was left to the con-sumer and the focus was on the proportion ofvacations spent in any one destination. Hencethe size of the destination was not consideredimportant (country versus village) but towhat extent the tourists felt they hadreturned back to a destination they had vis-ited in the past. This approach would allowfor the multi-dimensional nature of a holidayexperience and accommodate the telescopicdefinitions offered for a ‘destination’. Inessence the perceived ‘sameness’ of past holi-day destinations (and the proportion of pastholidays spent is just one destination) shouldbe the focus of the study rather than thenames of villages or countries, i.e. theresearch focus here is on the psychologicalpropensity for any individual to become loyaltowards a holiday destination. The destina-tion name, geographical boundaries andphysical attributes will be relevant when theresearch focus is on the reasons tourists doreturn to a specific destination. Such researchcan be used to enhance product (destina-tion) features and to create higher retentionrates. Furthermore, the reasons why touristsdo not want to return to a destination canhelp in designing quality improvements atthe destination.

The New Conceptualization of TourismDestination Loyalty

The concept of OSL postulates that each sub-ject has a natural, psychological tendency toeither seek or avoid experiences that generatestimulation. In tourism ‘language’ this meanseither variety-seeking or variety-avoiding holi-days. The argument here is that OSL can beused to identify those tourists who seek variedholidays (the opposite type of behaviour todestination loyalty). Once the variety-seekingtourists have been identified, the remainingholidaymakers are evaluated on the basis oftheir past holiday destination patterns and rel-

ative attitude towards repetitive holidays (theclassic loyalty concept). Therefore, the pro-posed conceptualization of destination loyaltydraws from the conceptual definition of con-sumer loyalty by Jacoby and Chestnut (1978,p. 80): ‘the biased (i.e. non-random), behav-ioural response, expressed over time, by somedecision making unit, with respect to one ormore alternative brands out of a set of brands,and it is a function of psychological (decisionmaking, evaluative) process’.

The tourist’s inherent destination loyaltyconceptual definition is, therefore, a biased(i.e. non-random) behavioural response,expressed over time, by an individual withrespect to one or more alternative holidaydestinations. It is a psychological tendencyand a function of the individual’s OSL andthe individual’s attitude towards repeated hol-idays. This attitude is an outcome of repeatedpersonal experience (and attitude activation)and therefore it is centrally held with confi-dence. The propensity to refer and intentionto return are indicators of this self-confi-dence. Figure 24.2 presents this tourist’sinherent destination loyalty concept.

The proposed inherent destination loyaltyconcept is based on the personality traits thatdirect the individual’s OSL. The desire forbalance in the stimulation levels will have animpact on the holiday decision making behav-iour. This behaviour is also guided by theindividual’s attitudes towards repetitive holi-days as well as their evaluation of their ownexpertise in travel decision making (confi-dence levels). The individual’s confidencelevel can be inferred from their propensity torecommend the destination to peers, makeindependent holiday arrangements, and beconfident to travel alone. This propositionhas support from past research since the‘Advocates’ at the top of the Loyalty Ladderare confident about their product evaluationand prepared to recommend the brand toothers as well. Furthermore, attitudes basedon personal (repetitive) past behaviour arealso argued to be more centrally held withmore confidence (Regan and Fazio, 1977;Smith and Swinyard, 1983; Allen et al., 1992).

Here it is important to recognize that overtbehaviour has been ‘filtered through’ situa-tional, intervening variables. In the destina-




Per

sona

lity

trai

tsB

arrie

rs to

vaca

tions

Hig

h re

peat

beha

viou

r

Low

rep

eat

beha

viou

r

Fee

dbac

k

Low

OS

L

Hig

hO

SL

Atti

tude

tow

ards

repe

titiv

eho

liday

s(+

, −)

Cen

tral

ly h

eld

conf

iden

ce in

ow

ntr

avel

dec

isio

n-m

akin

g an

dse

curit

y in

atti

tude

Atti

tude

tow

ards

repe

titiv

eho

liday

s(−

)

Cen

tral

ly h

eld

conf

iden

ce in

ow

ntr

avel

dec

isio

n-m

akin

g an

dse

curit

y in

atti

tude

Hig

hde

stin

atio

nlo

yalty

Late

ntde

stin

atio

nlo

yalty

No

vaca

tion

Low

dest

inat

ion

loya

lty

Spu

rious

dest

inat

ion

loya

lty

Fig.

24.

2.Th

e co

ncep

tual

izat

ion

of to

uris

t’s in

here

nt d

estin

atio

n lo

y alty

.


tion loyalty context, the typical barriers tovacations would be lack of finances, holidayentitlement, the wishes of the travel compan-ions, availability of vacation packages/flights,etc. This proposed triangulation of psycholog-ical measures (OSL, attitude towards repeti-tive holidays, indications of centrally heldconfidence) combined with the behaviouralmeasure of past travel career will enable theidentification of the high, low, latent and spu-rious destination loyal tourists. It is importantto recognize the group of people who choosenot to go on a vacation at all but prefer to stayat home, because they ‘like staying at home’.To some extent this could be an indication of‘home-destination loyalty’ and may create biasin the Low Destination Loyalty category unlessthis group is identified. Finally, it is alsoimportant to acknowledge the feedback frompast personal decisions. Depending on theperceived success of the holiday decision, suchfeedback will have an impact on the strengthof the attitude towards repetitive holidays aswell as the individual’s judgement of theirability to choose successful holidays.

Oppermann (2000) commented on howimpractical loyalty conceptualization was forprimary research purposes. To what extent,

then, can this proposed conceptualization forinherent destination loyalty be ‘framed’ forprimary research? There are several estab-lished ‘pen and paper’ measures for the OSLconcept; the most widely cited instrumentsare the Arousal Seeking Tendency scale (ver-sions I and II) (Mehrabian, 1978) and theSensation Seeking Scale (versions I–VI)(Zuckerman et al., 1964). Furthermore, therehave been some attempts to create an attitu-dinal measure for destination loyalty. Theproportion of past holidays spent on just onedestination can be inferred from qualitativeinterviews as well as quantitative surveys andthe confidence level indicators can be investi-gated through propensity to refer, make inde-pendent travel arrangements and travel alone(Niininen, 2002).

To conclude, any future research into des-tination loyalty should include both psycho-logical and behavioural measures to establishthe respondent’s inherent loyalty propensity.A triangulation of these two measures withvariety-seeking instruments would also assistin identifying the spurious/latent/low/highloyal tourists. Furthermore, measures of self-confidence should also be included as a con-trol variable in such studies.


References

Allen, C.T., Machleit, K.A. and Schultz Kleine, S. (1992) A comparison of attitudes and emotions as predic-tors of behaviour at diverse levels of behavioural experience. Journal of Consumer Research 18, 493–504.

Arnett, J. (1994) Sensation seeking: a new conceptualization and a new scale. Personality and IndividualDifferences 16, 289–296.

Arnett, J. (1996) Sensation seeking, aggressiveness, and adolescent reckless behaviour. Personality andIndividual Differences 20, 693–702.

Backman, S.J. (1991) An investigation of the relationship between activity loyalty and perceived con-straints. Journal of Leisure Research 23, 332–344.

Backman, S.J. and Crompton, J.L. (1991) Differentiating between high, spurious, latent and low loyaltyparticipants in two leisure activities. Journal of Park and Recreation Administration 9, 1–17.

Baloglu, S. and Erickson, R.E. (1998) Destination loyalty and switching behaviour of travellers: a Markowanalysis. Tourism Analysis 2, 119–127.

Baumgartner, H. and Steenkamp, J.-B.E.M. (1994) An investigation into the construct validity of thearousal seeking tendency scale, Version II. Psychological Measurement 54, 993–1001.

Bello, D.C. and Etzel, M.J. (1985) The role of novelty in the pleasure travel experience. Journal of TravelResearch Summer, 20–26.

Bloemer, J., de Ruyter, K. and Wetzels, M. (1999) Linking perceived service quality and service loyalty: amulti-dimensional perspective. European Journal of Marketing 32, 1082–1106.

Charlton, P. (1973) A review of shop loyalty. Journal of the Market Research Society 15, 35–51.Cook, R.L. and McCleary, K.W. (1983) Redefining vacation distances in consumer minds. Marketing

Research 10, 31–34.



Crotts, J.C. (1993) Personality correlates of the novelty seeking drive. Journal of Hospitality and LeisureMarketing 1, 7–29.

Darnell, A.C. and Johnson, P.S. (2001) Repeat visits to attractions: a preliminary economic analysis. TourismManagement 22, 119–126.

Day, G.S. (1969) A two-dimensional concept of brand loyalty. Journal of Advertising 9, 29–35.Dekimpe, M.G., Steenkamp, J.-B.E.M., Mellens, M. and Vanden Abeele, P. (1997) Decline and variability in

brand loyalty. International Journal of Research in Marketing 14, 405–420.Denison, T. and Knox, S. (1992) Profiling the Promiscuous Shopper. AIR MILES Travel Promotions, Crawley.Dick, A.S. and Basu, K. (1994) Customer loyalty: toward an integrated conceptual framework. Journal of the

Academy of Marketing Science 22, 99–113.Dowling, G.R. and Uncles, M. (1997) Do customer loyalty programs really work? Sloan Management Review

Summer, 71–82.Driver, L. (1996) What is loyalty in customer loyalty – the issues for the 90’s. The Researcher 1, 2–5.East, R., Harris, P., Willson, G. and Hammond, K. (1995) Correlates of first-brand loyalty. Journal of

Marketing Management 11, 487–497.Etzel, M.J. and Wahlers, R.G. (1985) The use of requested promotional material by pleasure travellers.

Travel Research 23, 2–6.Evans, M.J., Moutinho, L. and van Raaij, W.F. (1996) Applied Consumer Behaviour. Addison-Wesley, Harlow.Exter, T. (1986) Looking for brand loyalty. American Demographics 4, 32–56.Fiske, D.W. and Maddi, S.R. (1961) Functions of Varied Experience. The Dorsey Press, Homewood, Illinois.Gitelson, R.J. and Crompton, J.L. (1984) Insights into the repeat vacation phenomenon. Annals of Tourism

Research 11, 199–217.Godbey, G. and Graefe, A. (1991) Repeat tourism, play and monetary spending. Annals of Tourism Research

18, 213–225.Gyte, D.M. and Phelps, A. (1989) Patterns of destination repeat business: British tourists in Mallorca,

Spain. Journal of Travel Research Summer, 24–28.Hitchcock, M., King, V.T. and Parnell, M.J.G. (1994) Tourism in South-East Asia. Routledge, London.Howard, D.R., Edginton, C.R. and Selin, S.W. (1988) Determinants of program loyalty. Journal of Park and

Recreation Administration 6, 41–51.Jacoby, J. and Chestnut, R.W. (1978) Brand Loyalty Measurement and Management. John Wiley & Sons,

Chichester.Jacoby, J. and Kyner, D.B. (1973) Brand loyalty vs. repeat purchasing behaviour. Journal of Marketing

Research X, 1–9.Jones, T.O. and Sasser, W.E. Jr (1995) Why satisfied customers defect. Harvard Business Review

November–December, 88–99.Kish, G.B. and Donnenwerth, G.V. (1969) Interests and stimulus seeking. Journal of Counselling Psychology

16, 551–556.Lee, T.-H. and Crompton, J. (1992) Measuring novelty seeking in tourism. Annals of Tourism Research 19,

732–751.Mazursky, D. (1989) Past experience and future tourism decisions. Annals of Tourism Reseach 16, 333–344.Mehrabian, A. (1978) Characteristic individual reactions to preferred and unpreferred environments.

Journal of Personality 46, 717–731.Menon, S. and Khan, B. (1995) The impact of context on variety seeking in product choices. Journal of

Consumer Research 22, 285–295.Niininen, O.I. (2002) A study of the propensity for loyalty in tourism. PhD thesis, The University of Surrey,

Guildford, UK.Olson, J.C. and Jacoby, J. (1971) Construct validation study of brand loyalty. Proceedings of the American

Psychological Association (APA), 79th Annual Conference, pp. 657–658.Oppermann, M. (1997) First-time and repeat visitors to New Zealand. Tourism Management 18, 177–181.Oppermann, M. (1998) Destination threshold potential and the law of repeat visitation. Journal of Travel

Research 37, 131–137.Oppermann, M. (1999) Predicting destination choice – a discussion of destination loyalty. Journal of

Vacation Marketing 5, 51–65.Oppermann, M. (2000) Tourism destination loyalty. Journal of Travel Research 39, 78–84.



Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1994) Reassessment of expectations as a comparison stan-dard in measuring service quality, implications for further research. Journal of Marketing 58, 111–124.

Pritchard, M.P. and Howard, D.R. (1997) The loyal traveller, examining a typology of service patronage.Journal of Travel Research Spring, 2–10.

Pyo, S., Song, J. and Chang, H. (1998) Implications of repeat visitor patterns: the Cheju Island case.Tourism Analysis 13, 181–187.

Raj, S.P. (1985) Striking a balance between brand ‘popularity’ and brand loyalty. Journal of Marketing 49,53–59.

Regan, D.T. and Fazio, R. (1977) On the consistency between attitudes and behaviour: look to the methodof attitude formation. Journal of Experimental Social Psychology 13, 28–45.

Reichheld, F.F. and Teal, T. (1996) The Loyalty Effect: the Hidden Force Behind Growth, Profits, and Lasting Value.Bain and Company Inc., Boston, Massachusetts.

Riley, M., Niininen, O., Szivas, E.E. and Willis, T. (2001) The case for process approaches in loyaltyresearch in tourism. International Journal of Tourism Research 3, 23–32.

Rivers, M-J., Toh, R.J. and Withiam, G. (1991) Frequent-guest programs: do they fly? Cornell Hotel andRestaurant Administration Quarterly 32, 46–52.

Rogers, R. (1979) Commentary on ‘the neglected variety drive’. Journal of Consumer Research 6, 88–91.Ross, G.F. (1994) The Psychology of Tourism. Hospitality Press, Melbourne.Rowley, J. and Dawes, J. (2000) Disloyalty: a closer look at non-loyals. Journal of Consumer Marketing 17,

538–549.Ryan, C. (1995) Learning about tourists from conversations: the over 55s in Majorca. Tourism Management

16, 207–215.Smith, R.E. and Swinyard, W.R. (1983) Attitude-behaviour consistency: the impact of product trial versus

advertising. Journal of Advertising Research XX, 257–267.Snepenger, D.J. (1987) Segmenting the vacation market by novelty-seeking role. Journal of Travel Research

26, 8–14.Snyder, D.R. (1991) Demographic correlates to loyalty in frequently purchased consumer services. Journal

of Professional Services Marketing 8, 45–55.Srinivasan, M. (1996) New insight into switching behaviour: marketers can now put a numerical value on

loyalty. Marketing Research 8, 27–33.Uncles, M.D. and Ehrenberg, A.S.C. (1990) Brand choice among older consumers. Journal of Advertising

Research August/September, 19–22.Uncles, M. and Laurent, G. (1997) Editorial. International Journal of Research in Marketing 14, 399–404.van Trijp, H., Hoyer, W. and Inman, J. (1996) Why switch? Product category level explanations for true

variety-seeking behaviour. Journal of Marketing Research XXXIII, 281–292.Venkatesan, M. (1973) Cognitive consistency and novelty seeking. In: Ward, S. and Robertson, T.S. (eds)

Consumer Behaviour: Theoretical Sources. Prentice Hall, Englewood Cliffs, New Jersey, pp. 354–384.Wahlers, R.G. and Etzel, M.J. (1985) Vacation preference as a manifestation of optimal stimulation and

lifestyle experience. Journal of Leisure Research 17, 283–295.Zuckerman, M., Kolin, E.A., Rrice, L. and Zoob, I. (1964) Development of a sensation-seeking scale.

Journal of Consulting Psychology 28,477–482.



Chapter twenty-fiveMeasuring Comparative Performance of VacationDestinations: Using Tourists’ Self-reported Judgements as anAlternative Approach

Metin KozakSchool of Tourism and Hotel Management, Mugla University, 48000 Mugla, Turkey

Abstract

This study was designed to test the use of a practical approach developed: (i) to measure external perfor-mance of two international tourist destinations (Turkey and Mallorca); and (ii) to investigate their com-petitive position not only against each other but also against other major self-selected destinations on thebasis of the tourists’ perceptions of various self-reported destination attributes. The sample populationchosen was comprised of British tourists who: (i) visited Turkey and Mallorca in the summer of 1998; and(ii) had travelled to other international destinations in the preceding 4 years. An open-ended (verbal)questionnaire was designed to achieve the research objectives. From the research findings, it is apparentthat the analysis of tourists’ self-reported judgements is relevant to report differences in the strengths andweaknesses of one destination vis-à-vis those of another in the same competitor set. Implications of thefindings for the theory of consumer behaviour in tourism, for understanding qualitative research, for des-tination management and marketing, and for future research are discussed.

Introduction

Tourist destinations are a key component ofthe tourism system. On the one hand, eachdestination offers a variety of products andservices to attract visitors. Conversely, eachvisitor has the opportunity and freedom tochoose where to go for vacation from a virtu-ally unlimited set of alternative destinations(Crompton, 1992). Different factors influ-ence the destination choice because eachconsumer may have different motivations and

preferences for different destinations. Moststudies of tourist destinations and consumerbehaviour have focused on the relationshipbetween attitude towards a specific destina-tion, the image of that destination, touristpreferences in destination choice, and/or vis-itor satisfaction at the destination.

As with other industries, tourist destina-tions compete with each other (Heath andWall, 1992). Developments in internationaltourism and transport technology have inten-sified the competition between international



tourist destinations. New destinations becomeestablished, some existing ones grow, othersdecline. Many tourists have experience atother destinations. It is, therefore, expectedthat visitors will make comparisons betweenfacilities, attractions and service standards(Laws, 1995). It is argued that a visitor selectsa destination from alternatives and evaluateseach alternative considering its potential todeliver the benefits sought (Mayo and Jarvis,1981). As a result, factors such as natural, cul-tural and historical resources, infrastructure,accessibility, attractions and facilities areaccepted as elements of tourist destinationcompetitiveness.

The concepts of competitiveness and per-formance improvement are also interrelated(Zairi, 1996). Improved performanceenhances competitive advantage. Conversely,poor performance must be corrected beforethe destination can effectively compete withothers. Based upon the related literature(Melcher et al., 1990), it seems that measur-ing performance, as a key issue in competi-tiveness, could help destination managersidentify areas where their destination per-forms well and poorly and, thereby, focusattention on areas of needed improvement.

As each destination may have its ownadmirers, tourists satisfied in one destinationmay differ from those satisfied at other desti-nations. Furthermore, how they perceive theattributes of one destination may differ fromhow they perceive the attributes of other des-tinations. This creates serious problems in themeasurement of external performance.Nevertheless, previous destination compari-son and competitiveness research typicallyemploy a single attribute rating scale (Javalgiet al., 1992; Driscoll et al., 1994). Particularlywhen examining multiple destinations anddifferent consumer segments, such scales maynot measure the appropriate attributes.Furthermore, they typically do not measurecompetitive superiority or inferiority.Therefore, the proposed models in the litera-ture may not help to evaluate a destination’scomparative service performance. A moreeffective comparative analysis of destinationperformance may help to better understandconsumer behaviour and also assist destina-tion authorities to establish effective position-

ing strategies and explore their core compe-tencies for each market.

In the broader context of destination com-petitiveness research, a destination needs tobe compared with more than two destina-tions. In so doing, the destination managersmay be able to understand the level of theirown performance not only against one spe-cific destination but also against their majorcompetitors. In this research, it is expectedthat sample populations have direct experi-ence in order to respond accurately to ques-tions regarding their actual holidayexperiences in each destination. Otherwise,findings do not accurately reflect the perfor-mance of destinations on specific attributes.Thus, there is a need for a better understand-ing of tourists’ judgements in measuring com-parative destination performance.

The literature suggests that open-endedquestions can be used to collect data regard-ing both negative and positive customer com-ments (Danaher and Haddrell, 1996). Thesedata can then be compared with customers’overall evaluations of the service or destina-tion. Open-ended questionnaires have beenused to examine tourists’ positive and nega-tive experiences (Pearce and Caltabiano,1983; Johns and Lee-Ross, 1995; Jackson et al.,1996), destination images (Reilly, 1990), pur-chase–consumption systems (King andWoodside, 2001), and antecedents and conse-quences of tourist satisfaction (Decrop,2001). In these studies, the questionnaireswere distributed to allow respondents to replyin their own words, but not in an attempt at adirect comparison with other tourist establish-ments or destinations.

Little has been done to examine the com-parative performance of different interna-tional tourist destinations. Some researchershave compared one destination’s perfor-mance with other similar and competing des-tinations on the basis of several pre-identifiedgeneric variables (Goodrich, 1977, 1978;Haahti and Yavas, 1983; Haahti, 1986;Pearce, 1997). A review of the related litera-ture indicates that the respondents were notgiven any freedom to report what they foundto be good or poor at the study destinationson the basis of self-selected items beingreflected in their own words. Moreover, with

286 M. Kozak


limited exceptions (e.g. King, 1994; Kozakand Rimmington, 1999; Kozak, 2003), muchof the research conducted using primarymethods was undertaken without evidencethat respondents had actually been to thestudy destinations. Therefore, the existingresearch does not provide a full account ofdestination performance and competitive-ness (see Table 25.1 for the list of the relatedstudies and their details).

As a part of the service industry, tourismdiffers from other industries in that itrequires customers (or tourists) to participatedirectly both in the production and consump-tion of holiday products and services. Thishighlights the importance of measuring satis-faction and image perceptions of tourists whohave actually experienced the performanceof organizations or tourist destinations. Inother words, it is not reasonable in tourismresearch to avoid the experiences or the feed-back of actual customers by asking outsidersabout their ideas or feelings instead.

Taking these issues into consideration, thepurpose of this study was to test the use of aresearch methodology developed: (i) to mea-sure external performance of two interna-tional tourist destinations (Turkey andMallorca); and (ii) to investigate their com-petitive position not only against each other,but also against other major internationaldestinations on the basis of several self-selected attributes obtained from the analysisof an open-ended questionnaire. The samplepopulation was comprised of British touristswho: (i) visited Turkey and Mallorca in thesummer of 1998; and (ii) had travelled toother international destinations in the pre-ceding 4 years.

Questionnaire Design

An open-ended questionnaire was designedto accomplish the study objectives. Theinstrument required that tourists visitingTurkey and Mallorca had separately been toother international destinations (at leastonce) in the previous 4 years. The time waslimited to the beginning of 1995 (the last 4years), since the respondents might have haddifficulty in recalling their earlier experi-

ences. A draft questionnaire was pilotedamong ten people working in a British uni-versity. The pilot study sample was chosenfrom the university directory and a copy ofthe questionnaire was sent via the university-based internal mail system. Subsequently,minor modifications were incorporated intothe questionnaire and the final draft was pro-duced and copied. The questionnaire instru-ment consisted of six open-ended questions.

The first question identified other inter-national destinations the respondents hadvisited since the beginning of 1995. The sec-ond question gave the respondents freedomto choose any one of those destinations tocompare with their current holiday. Thethird question asked for the name of the des-tination where the respondents spent theirholidays in Turkey and Mallorca. The fourthquestion gave an opportunity to understandwhat attributes of their current trip were bet-ter than the selected comparison destination(positive aspects of tourism). Similarly, thefifth question identified the potential weak-nesses that Turkey and Mallorca need toimprove in order to increase their competi-tive position against the comparison destina-tion (negative aspects of tourism). In thefinal question, the respondents wererequested to state the name of the destina-tion which they liked best and the underlyingreason(s). This could be helpful in identify-ing destinations to which the respondentsmay return in the future and reasons that areof importance in attracting tourism. In otherwords, these findings may indicate thestrengths of destinations with which therespondents were most satisfied.

Data Collection

The study sample consisted of British touristsvisiting Turkey and Mallorca. Surveys andobservations were restricted to a 3-weekperiod in each country during the peak sea-son in the summer of 1998. Passengers wereapproached in airport departure lounges andasked which country they were from. Thosefrom the UK were then asked if they wouldlike to participate in the survey. As a result ofthe nature of the information sought (per-

Measuring Comparative Performance of Vacation Destinations 287


288 M. Kozak

Table 25.1. Overview of previous destination comparison/competitiveness research.

Author Criteria Explanation

Goodrich (1977) Tourist perceptions of similarities anddifferences between nine regions onwater sports and sports, historical andcultural interests, scenic beauty,hospitality, rest and relaxation,shopping facilities, cuisine,entertainment and accommodation

Goodrich (1978) Tourist perceptions of nine regionsand their intention to choose them. There is no evidence that theAttributes were same as above respondents had actually

Haahti and Yavas (1983); Tourist perceptions of 12 European been in person to all sampleHaahti (1986) countries on value for money, destinations under

accessibility, sports facilities and investigation (indirectother activities, nightlife and measure of destinationentertainment, peace and quietness, performance). The samplehospitality, wilderness, tracking and population is not allowed tocamping, cultural experience, scenery, self-report the attributeschange from the usual destinations perceived to be important and

Calantone et al. (1989) Tourist perceptions of several the destinations to bedestination attributes (including compared againstshopping facilities, hospitality, safety, food, culture, tourist attractions, touristfacilities, nightlife and entertainment,scenery, beaches and water sports)

Javalgi et al. (1992) Travellers’ perceptions of Europeandestinations (as four major regions)about 27 attributes

Driscoll et al. (1994) Tourist perceptions of 12 destinationson 18 attributes such as facilities,landscape, safety, climate, culture,modern society, different experience,value for money, accessibility, shoppingfacilities, organized activities,cleanliness, family oriented, exoticplace, outdoor activities, religious values,hospitality, nightlife and entertainment

Faulkner et al. (1999) Analysis of travel agents’ perceptionsof core tourist attractions

Botho et al. (1999) Tourist motivations and tourist per-ceptions of entertainment, infrastructure,physical environment and wildlife

Kozak and Rimmington (1999) Tourist satisfaction with several The study provides sufficientattributes, tourist complaints, holiday evidence that the respondentstaking behaviour with respect to had actually been in person todestinations, years and seasons all sample destinations under

investigation (direct measureof destination performance).The sample population isallowed to self-report theattributes perceived to beimportant and the destinationsto be compared against

Source: Kozak (2003).


sonal opinions), passengers completed thequestionnaires themselves. Those who stayedin private accommodation, with relatives orfriends or were on cruises were not given thequestionnaire. It is important to note that,with the purpose of obtaining different viewsand avoiding repetition and imitation, thequestionnaire was delivered to only one per-son in each family or group. Given that thelength of the holiday may influence destina-tion perceptions, only tourists who had stayedat least 1 week on holiday were included inthe survey. All questionnaires, whether com-pleted or not, were returned before passen-gers embarked.

Table 25.2 reports the distribution of ques-tionnaires by study site. At Palma Airport(Mallorca), the total number of question-naires completed was 198, of which four ques-tionnaires were unusable. At DalamanAirport (Turkey), the total number of ques-tionnaires completed was 198, of which 18were unusable. Those questionnaires elimi-nated were incomplete. Refusals in the tablerefer to those who either were not suitable forthe survey or refused to participate.

Analysis of Findings

Content analysis was employed to analyse thedata derived from the open-ended question-naire (Robson, 1993), resulting in lists ofwords (or items) for each question. Theseitems were ordered according to the numberof times they appeared. Percentage valueswere then calculated for each item by divid-ing the frequency counts by the sample sizein each tourist group. The higher the fre-

quency value, the better the item was consid-ered by respondents, for the question wasdesigned to demonstrate how likely the desti-nation was to be perceived to be better thanother destinations. In contrast, the higher thepercentage value, the worse the item, for thequestion designed to demonstrate how likelythe destination was to be perceived to beworse than other destinations. Also, somedirect quotations from the open-ended ques-tionnaire were inserted into appropriatepoints to emphasize key differences.

Each reply was numbered and summarizedinto a list of keywords and phrases, whichencapsulated the customers’ experiences.Three sets of cards were created, one for theitems relating to experience of better services(satisfiers), one for the items describing worseservices (dissatisfiers) and one for the itemsdescribing what made the holiday satisfactoryand affecting which destination to revisit(Tables 25.3–25.9). The respondents weremore likely to choose a Mediterranean desti-nation to compare with their holiday experi-ences either in Mallorca or Turkey. This hashelped to make a successful and reliable com-parison between similar destinations.

Identifying Strengths of Mallorca andTurkey

The first seven highly ranked attributes foundby British tourists visiting Turkey to be betterthan those of other destinations visited sincethe beginning of 1995 were positive attitudesof the local people and staff towards tourists(hospitality) (52%), level of prices (34%),weather (17%), plenty to do and see (15%),overall cleanliness of the resorts (13%),scenery (11%), and quality and variety offood (11%), respectively (Table 25.3). Overallcleanliness of the resorts (19%), level ofprices (18%), hospitality (18%), cleanlinessand quality of beaches and the sea (16%),weather (12%), plenty to do and see (12%),and the availability and suitability of enter-tainment and nightlife (10%) were, respec-tively, the most significant attributes ofMallorca (Table 25.4). When the findingsabout Turkey and Mallorca are compared,the overall cleanliness, level of prices,


Table 25.2. Distribution of questionnaire bystudy site.

Process Palma Airport Dalaman Airport

Approached 352 322N/A (or refusals) 154 124Delivered 198 198Returned 198 184Eliminated 4 18Analysed 194 166


290 M. Kozak

Tab

le 2

5.3.

Attr

ibut

es fo

und

to b

e be

tter

in T

urke

y th

an in

com

petin

g de

stin

atio

ns (

n=

166

).

Gra

nTo

tal

Can

ada

Item

sG

reec

eS

pain

Can

aria

Por

tuga

lB

alea

ricM

alta

Fra

nce

Italy

Cyp

rus

Tuni

sia

Egy

ptM

ed.

% M

ed.

and

US

AO

ther

sTo

tal

%

Hos

pita

lity

289

53

113

23

51

272

43.3

212

8651

.8Le

vel o

f pric

es8

412

47

12

31

01

4325

.92

1257

34.3

Wea

ther

86

44

10

10

10

025

15.0

22

2917

.4M

ore

to d

o an

d se

e6

35

23

10

02

01

2313

.80

225

15.0

Ove

rall

clea

nlin

ess

42

50

30

00

11

117

10.2

04

2112

.6S

cene

ry1

63

13

20

01

00

1710

.20

219

11.4

Foo

d7

40

13

20

00

00

1710

.20

219

11.4

Val

ue fo

r m

oney

42

30

11

10

10

013

7.8

10

148.

4N

ight

life

and

ente

rtai

nmen

t3

20

00

02

00

01

84.

81

09

5.4

Not

com

mer

cial

ized

20

21

20

00

00

18

4.8

00

84.

8A

ccom

mod

atio

nfa

cilit

ies

30

10

10

00

10

05

3.0

02

84.

8Q

uiet

1–

12

30

00

00

07

4.2

00

74.

2C

lean

lines

s of

beac

hes

and

sea

11

22

10

00

00

07

4.2

00

74.

2C

ultu

re0

22

00

11

01

00

74.

20

07

4.2

Sho

ppin

g fa

cilit

ies

22

00

00

00

00

04

2.4

10

53.

0R

esta

uran

ts1

00

02

00

00

00

31.

81

04

2.4

Saf

ety

and

secu

rity

00

10

00

00

00

12

1.2

02

42.

4Tr

ansp

ort s

ervi

ces

40

00

00

00

00

04

2.4

00

42.

4A

tmos

pher

e0

02

00

01

00

00

31.

80

14

2.4

Ove

rall

serv

ices

inth

e re

sort

00

00

00

11

00

02

1.2

01

31.

8A

cces

s to

UK

00

00

00

00

00

00

00

33

1.8

Sta

ndar

d of

faci

litie

s0

10

10

01

00

00

31.

80

03

1.8

Wat

er s

uppl

y1

00

00

00

00

00

10.

60

12

1.2

Lang

uage

/co

mm

unic

atio

n1

01

00

00

00

00

21.

20

02

1.2

Wat

ersp

orts

20

00

00

00

00

02

1.2

00

21.

2La

rge

reso

rts

00

00

00

00

00

01

0.6

00

10.

6



Tab

le 2

5.4.

Attr

ibut

es fo

und

to b

e be

tter

in M

allo

rca

than

com

petin

g de

stin

atio

ns (

n =

194

) (s

ame

com

men

t).

Gra

nTo

tal

%Ite

ms

Gre

ece

Spa

inC

anar

iaP

ortu

gal

Turk

eyM

alta

Fra

nce

Italy

Cyp

rus

Tuni

sia

Men

orca

Ibiz

aM

ed.

Med

.U

SA

Oth

erTo

tal

%

Ove

rall

clea

nlin

ess

73

63

61

00

12

33

3518

.00

237

19.0

Leve

l of p

rices

42

13

11

61

30

10

2311

.86

736

18.5

Hos

pita

lity

(peo

ple)

25

54

40

50

12

32

3317

.00

235

18.0

Bea

ches

and

sea

32

75

30

10

20

14

2814

.41

231

15.9

Wea

ther

20

47

00

31

00

01

189.

22

525

12.8

Mor

e to

do

and

see

70

51

10

01

00

42

2110

.80

223

11.8

Foo

d3

11

10

26

11

01

320

10.3

02

2211

.3N

ight

life

and

ente

rtai

nmen

t1

03

31

20

01

12

115

7.7

23

2010

.3S

cene

ry3

12

20

10

03

01

013

6.7

30

168.

2F

amily

res

ort

13

31

00

00

10

02

115.

61

214

7.2

Loca

l tra

nspo

rt2

20

22

10

00

00

09

4.6

03

126.

1S

tand

ard

of fa

cilit

ies

20

13

10

00

03

00

105.

12

012

6.1

Acc

omm

odat

ion

faci

litie

s0

02

31

00

00

12

312

6.1

00

126.

1A

cces

s to

UK

00

20

00

00

10

00

31.

53

410

5.1

Sho

ppin

g fa

cilit

ies

10

34

01

00

01

00

105.

10

010

5.1

Res

taur

ants

02

11

00

10

01

00

63.

03

09

4.6

Res

ort a

irpor

t3

21

10

10

00

10

09

4.6

00

94.

6F

acili

ties

for

child

ren

02

11

00

10

20

10

84.

10

08

4.1

Atm

osph

ere

10

11

00

20

00

00

52.

52

07

3.6

Qui

et1

20

10

00

00

00

26

3.0

00

63.

0La

ngua

ge/

com

mun

icat

ion

01

12

00

01

10

00

63.

00

06

3.0

Larg

e re

sort

00

01

00

00

00

40

52.

50

05

2.5

Sm

all r

esor

t0

01

10

01

00

00

03

1.5

10

42.

0Le

ss c

omm

erci

aliz

ed0

11

10

00

00

00

03

1.5

00

31.

5S

afet

y an

d se

curit

y0

00

00

00

00

10

01

0.5

20

31.

5R

oad

and

traf

ficco

nditi

ons

00

02

01

00

00

00

31.

50

03

1.5


292 M. Kozak

Tab

le 2

5.5.

Attr

ibut

es to

be

cons

ider

ed w

hile

cho

osin

g th

e be

st d

estin

atio

n (T

urke

y, n

= 1

66).

Item

sTu

rkey

Gre

ece

Cyp

rus

Spa

inG

ran

Can

aria

Tuni

sia

Mal

taTo

tal M

ed%

Med

.U

SA

Oth

ers

Tota

l%

Hos

pita

lity

(peo

ple)

620

10

00

063

37.9

11

6539

.1M

ore

to d

o an

d se

e22

00

10

00

2313

.81

024

14.4

Wea

ther

140

01

00

015

9.0

10

169.

6Le

vel o

f pric

es15

00

00

00

159.

00

015

9.0

Ove

rall

clea

nlin

ess

80

00

11

010

6.0

31

148.

4V

alue

for

mon

ey12

00

00

00

127.

11

114

8.4

No

hass

le1

41

22

10

116.

61

113

7.8

Nig

htlif

e an

d en

tert

ainm

ent

100

00

00

111

6.6

10

127.

2N

ot c

omm

erci

aliz

ed6

20

00

00

84.

80

210

6.0

Acc

omm

odat

ion

faci

litie

s7

00

00

10

84.

80

210

6.0

Sce

nery

90

00

00

09

5.4

00

95.

4F

ood

80

00

10

09

5.4

00

95.

4C

ultu

re a

nd h

isto

ry6

10

00

00

74.

20

07

4.2

Bea

ches

and

sea

21

01

00

04

2.4

00

42.

4S

afet

y an

d se

curit

y3

00

00

00

31.

80

03

1.8

Lang

uage

/com

mun

icat

ion

10

00

00

12

1.2

10

31.

8S

tand

ard

of fa

cilit

ies

30

00

00

03

1.8

00

31.

8Q

uiet

20

00

00

02

1.2

00

21.

2A

tmos

pher

e1

00

00

00

10.

60

12

1.2



Tab

le 2

5.6.

Attr

ibut

es to

be

cons

ider

ed w

hile

cho

osin

g th

e be

st d

estin

atio

n (M

allo

rca,

n =

194)

.

Gra

nIte

ms

Mal

lorc

aTu

nisi

aTu

rkey

Spa

inG

reec

eM

alta

Ibiz

aM

enor

caC

anar

iaC

ypru

sP

ortu

gal

Fra

nce

Italy

Med

.%

Med

. US

AO

ther

sTo

tal

%

Hos

pita

lity

(peo

ple)

221

02

00

11

20

00

029

14.9

30

3216

.4M

ore

to d

o an

d se

e20

00

20

01

11

11

10

2814

.42

030

15.4

Bea

ches

and

sea

190

01

10

01

01

00

023

11.8

00

2311

.8Le

vel o

f pric

es13

00

20

01

11

00

00

189.

21

221

10.8

Fam

ily o

rient

ed18

00

01

00

00

00

00

199.

72

021

10.8

Acc

omm

odat

ion

faci

litie

s16

00

00

00

00

10

00

178.

71

119

9.7

Ove

rall

clea

nlin

ess

130

02

00

02

00

00

017

8.7

01

189.

7N

o or

less

com

mer

cial

ized

21

00

11

40

00

20

011

5.6

03

147.

2R

elax

ed a

tmos

pher

e8

00

20

00

01

00

00

115.

60

011

5.6

Sta

ndar

d of

faci

litie

s8

01

00

00

00

00

00

94.

60

09

4.6

Qui

et1

00

11

01

10

01

00

63.

01

18

4.1

Foo

d5

00

00

00

00

11

00

73.

61

08

4.1

Fac

ilitie

s fo

r ch

ildre

n3

00

20

00

00

01

00

63.

01

07

3.6

Wea

ther

40

00

00

00

10

01

06

3.0

01

73.

6A

cces

s to

UK

50

00

00

00

00

00

05

2.5

50

52.

5C

ultu

re2

00

00

00

00

00

10

31.

50

25

2.5

Ava

ilabl

e fo

r al

l age

s5

00

00

00

00

00

00

52.

50

05

2.5

Saf

ety

and

secu

rity

40

00

00

00

00

00

04

2.0

00

42.

0S

hopp

ing

faci

litie

s2

00

00

01

01

00

00

42.

00

04

2.0

Res

taur

ants

40

00

00

00

00

00

04

2.0

00

42.

0B

ritis

h or

ient

ed3

00

00

00

01

00

00

42.

00

04

2.0

Nig

htlif

e an

den

tert

ainm

ent

10

00

00

10

00

00

02

1.0

01

31.

5S

ervi

ces

20

00

00

00

00

00

02

1.0

10

31.

5S

mal

l res

orts

10

00

00

11

00

00

03

1.5

00

31.

5S

cene

ry1

00

00

00

00

00

01

21.

00

02

1.0

Loca

l tra

nspo

rt2

00

00

00

00

00

00

21.

00

02

1.0

Lang

uage

/co

mm

unic

atio

n0

00

00

00

00

11

00

21.

00

02

1.0


294 M. Kozak

Tab

le 2

5.7.

Attr

ibut

es fo

und

to b

e w

orse

in T

urke

y th

an in

com

petin

g de

stin

atio

ns (

n =

166

).

Gra

nTo

tal

Can

ada

Item

sG

reec

eS

pain

Can

aria

Por

tuga

lB

alea

ricM

alta

Fra

nce

Italy

Cyp

rus

Tuni

sia

Egy

ptM

ed.

% M

ed.

and

US

AO

ther

sTo

tal

%

Has

sle

136

144

52

21

11

251

30.7

24

5734

.3R

oad

and

traf

fic c

ondi

tions

62

41

10

01

10

016

9.6

02

1810

.8O

vera

ll cl

eanl

ines

s2

25

10

00

00

10

116.

60

314

8.4

Cle

anlin

ess

of b

each

esan

d se

a3

30

11

02

00

00

106.

00

111

6.6

Air-

cond

ition

ing

02

20

00

01

10

06

3.6

02

84.

8W

eath

er0

22

02

00

00

00

63.

60

17

4.2

Res

ort a

irpor

t2

00

01

00

00

00

31.

81

37

4.2

Noi

se a

nd c

row

d0

00

10

20

10

00

42.

40

26

3.6

Com

mer

cial

ized

40

10

00

00

00

05

3.0

01

63.

6A

ccom

mod

atio

n fa

cilit

ies

10

30

00

00

00

04

2.4

01

53.

0To

ilet f

acili

ties

02

01

11

00

00

05

3.0

00

53.

0F

ood

00

20

01

00

00

03

1.8

00

31.

8Le

ss to

do

and

see

00

00

00

00

10

01

0.6

11

31.

8W

ater

qua

lity

00

00

10

01

00

02

1.2

01

31.

8Le

vel o

f pric

es0

00

00

00

00

00

00

02

21.

2N

othi

ng fo

r ch

ildre

n0

11

00

00

00

00

21.

20

02

1.2

Unt

idy

10

00

10

00

00

02

1.2

00

21.

2W

ater

spor

ts0

01

10

00

00

00

21.

20

02

1.2

Saf

ety

and

secu

rity

00

10

00

00

00

01

0.6

00

10.

6Lo

ng fl

ight

01

00

00

00

00

01

0.6

00

10.

6S

ignp

ostin

g ov

eral

l0

00

00

00

10

00

10.

60

01

0.6

Sta

ndar

d of

faci

litie

s0

01

00

00

00

00

10.

60

01

0.6

Atti

tude

tow

ards

wom

en1

00

00

00

00

00

10.

60

01

0.6



Tab

le 2

5.8.

Attr

ibut

es fo

und

to b

e w

orse

in M

allo

rca

than

in c

ompe

ting

dest

inat

ions

(n

=19

4).

Gra

nTo

tal

%Ite

ms

Gre

ece

Spa

inC

anar

iaP

ortu

gal

Turk

eyM

alta

Fra

nce

Italy

Cyp

rus

Tuni

sia

Men

orca

Ibiz

aM

ed.

Med

.U

SA

Oth

erTo

tal

%

Com

mer

cial

ized

51

32

10

42

11

43

2713

.94

435

18.0

Ove

rcro

wde

d2

11

40

13

10

12

117

8.7

51

2311

.8N

oisy

12

32

11

01

01

11

147.

23

219

9.7

Leve

l of p

rices

11

44

00

10

10

20

147.

20

216

8.2

Dirt

y (c

lean

lines

s)1

20

10

00

02

03

09

4.6

23

147.

2N

ight

life

and

ente

rtai

nmen

t0

14

11

00

00

00

18

4.1

22

126.

1F

ood

21

02

10

10

00

01

84.

11

211

5.6

Hos

pita

lity

(peo

ple)

10

30

10

10

01

10

84.

10

210

5.1

Roa

d an

d tr

affic

cond

ition

s0

10

11

11

01

01

07

3.6

10

84.

1S

hopp

ing

faci

litie

s0

10

10

00

10

00

25

2.5

10

63.

0B

each

es a

nd s

ea2

00

00

00

02

00

04

2.0

10

52.

5A

ccom

mod

atio

nfa

cilit

ies

00

10

00

00

01

00

21.

02

15

2.5

Loca

l tra

nspo

rt0

00

01

00

00

01

02

1.0

11

42.

0A

ir-co

nditi

onin

g0

00

01

10

01

00

03

1.5

01

42.

0W

eath

er1

01

01

00

00

00

03

1.5

00

31.

5To

ilet f

acili

ties

01

11

00

00

00

00

31.

50

03

1.5


296 M. Kozak

Table 25.9. Competitiveness set of Turkey and Mallorca.

Mallorca (n =194) Turkey (n =166)

Destinations n % Rank n % Rank

Canary Islands 54 27.8 1 43 25.9 2USA 49 25.2 2 30 18.0 4Greece 48 24.7 3 44 26.5 1France 44 22.6 4 19 11.4 5Spain 42 21.6 5 40 24.0 3Ibiza 32 16.4 6 13 7.8 6Menorca 27 13.9 7 5 3.0 17Portugal 23 11.8 8 11 6.6 9Cyprus 22 11.3 9 12 7.2 8Italy 19 9.7 10 13 7.8 7Turkey (Mallorca) 17 8.7 11 8 4.8 11Malta 12 6.1 12 11 6.6 10Caribbean 12 6.1 13 6 3.6 16Tunisia 11 5.6 14 4 2.4 19Austria 6 3.0 15 5 3.0 18Germany 5 2.5 16 2 1.2 25Canada 4 2.0 17 6 3.6 15Mexico 4 2.0 18 4 2.4 20Bulgaria 3 1.5 19 6 3.6 14Belgium 3 1.5 20 3 1.8 22Holland 3 1.5 21 7 4.2 13Sweden 2 1.0 22 0 0 0Denmark 2 1.0 23 0 0 0Switzerland 2 1.0 24 0 0 0Israel 2 1.0 25 2 1.2 27Hong Kong 2 1.0 26 0 0 0Thailand 2 1.0 27 4 2.4 21Australia 1 0.5 28 8 4.8 12Indonesia 1 0.5 29 2 1.2 28Kenya 1 0.5 30 2 1.2 29Egypt 1 0.5 31 3 1.8 24Czech Republic 1 0.5 32 3 1.8 23South Africa 1 0.5 33 2 1.2 30Morocco 1 0.5 34 0 0 0India 0 0 0 3 1.8 25China 0 0 0 2 1.2 31Maldives 0 0 0 2 1.2 32

As only one respondent from each sample destination had been to Singapore,Malaysia, Norway, Romania, Andorra, Sri Lanka, Mauritius, Bangladesh, Fiji,Gambia, Zambia and Tanzania, these places were excluded from the analysis of set.

hospitality, weather, and plenty to do and seewere the items found to be similar in favourof Turkey and Mallorca in comparison withother destinations. Although hospitality andlevel of prices were joint attributes of bothdestinations, their rankings for Turkey hadhigher percentage values than for Mallorca.Therefore, in addition to these two attributes,

scenery, and quality and variety of food werethose favouring Turkey, whereas quality andcleanliness of beaches and the sea, and suit-ability of entertainment and nightlifefavoured Mallorca.

The analysis revealed a significant gapbetween tourist perceptions of the price levelsin competing destinations and Turkey. The


latter was found to be more satisfactory andmuch cheaper. Turkey is perceived as an inex-pensive, sun and sea destination and hasbecome increasingly popular in the last 15years among tourists who travel to enjoy goodweather, sea and beaches. As a result of thepresent author’s personal observations, thelevel of prices in Mallorca seems to be higher,particularly for food and drink, gifts, visitingattractions and day trips, although it variesbetween resorts and between serviceproviders. The Turkish Lira (TL) is losing itsvalue against British Sterling and the GermanMark. The TL lost value against Sterling by115% and against DM by 106% between 1998and 2000. One repeat tourist already explainsthe potential reason for choosing Turkey inthe future as ‘the strong British Pound makesit a good value holiday’.

Good weather, found to be an importantelement of both places, is the main motiva-tion for north European people to take theirsummer vacations in the Mediterranean. It isclear that people travel to Turkey andMallorca to enjoy good weather. Several rea-sons can be speculated for this result, e.g. torelax, to be emotionally and physicallyrefreshed, and to engage in summer sportactivities. The associated products inMallorca and Turkey are yachting, beaches,sports (such as golf, mountain biking, hik-ing), watersports and cruising. Turkey wasalso perceived as having better facilities andactivities. Turkey offers its guests various hol-iday opportunities. There are daily or 2–3-day tours to main cultural and historicalattractions, as well as daily boat tours.Resorts in Turkey are also richer in water-sports and shopping facilities. ‘I enjoyedMarmaris (Turkey) the best because thereare many activities you can participate inboth at the hotels and on the beach. It isalso a bigger town; so there is lots more tosee…’. Although the availability of facilitiesand services is in Turkey’s favour, there isone significant complaint, the need toimprove air-conditioning services, includingthe accommodation facilities, because it isvery hot in summer time.

There was only one variable found to bebetter in Mallorca – the availability of facili-ties and activities for children. This might

be a significant issue for those with childrenchoosing Mallorca. As one tourist pointsout, ‘Mallorca has greatly improved for afamily holiday since our last visit in 1984.Mallorca is a good option for a family holi-day’. As observations indicate, Mallorca pro-vides a number of facilities and activitiesparticularly for family groups, such assports, watersports, private swimming pools,nursery services and playgrounds. Somerestaurants in the area attempt to attractthis market segment by offering a special,half-price menu for children under 12 andproviding a private playground. They alsoserve a special menu for those who areunder 12 years old. In Turkey, such servicesare still in their infancy and are providedonly by some large establishments like holi-day villages and five-star hotels. Most impor-tantly, the infrastructure in many Turkishresorts (e.g. Marmaris) is not designed forparents to be able to push their pushchairson the pedestrian roads.

Although the respondents had a highertendency to rate either Mallorca or Turkeyas the best destination in terms of their holi-day experiences, the findings could be sig-nificant in understanding which factors(attributes) were effective in their decisionmaking process and why Mallorca or Turkeywas selected as the best. Hospitality (39%),plenty to do and see (14%), weather (10%),level of prices (9%), overall cleanliness(8%), and value for money (7%) were attrib-utes the respondents liked most and consid-ered to be important when choosing Turkeyas the best compared to other destinations(Table 25.5). Moreover, hospitality (16%),plenty to do and see (15%), quality ofbeaches and sea (12%), level of prices(11%), being family oriented (11%), overallcleanliness (10%), and suitability of accom-modation facilities (10%) were the similarattributes of Mallorca (Table 25.6). Exceptfor hospitality and plenty to do and see, thelisted attributes vary between Turkey andMallorca. There is ground to argue that hos-pitality of local people and plenty to do andsee while on vacation are more importantattributes than ever when evaluating tourists’overall experiences in respect to their holi-days at these sample destinations.



Identifying Weaknesses of Mallorca andTurkey

Among the attributes found to be worse inTurkey than in other self-selected destina-tions were harassment (the perceived nega-tive attitudes of local shopkeepers in theresort to sell goods and services) (34%), traf-fic (poor conditions of roads and driving)(11%), overall dirtiness of the destinations(8%), poor quality and dirtiness of beachesand the sea (7%), and the lack of air-condi-tioning systems (5%) (Table 25.7). Thehighly ranked negative attributes of Mallorcawere over-commercialization (18%), over-crowding (12%), noise (10%), level of prices(8%), dirtiness (7%), lack of entertainmentand nightlife (6%), and poor quality and vari-ety of food (6%) (Table 25.8). As can be seen,except for over-commercialization, the twolists of negative comments were the same inMallorca and Turkey. This may be a result ofmovements in tourism to commercialize theindustry. The main problem in Turkey wasthat respondents felt pressured by local shop-keepers to buy something from their shops.The other elements of the list were mainlybased on tangible items (e.g. poor traffic,dirtiness, and lack of air-conditioning). Thecomplaints presented by those in Mallorcawere mostly related to the results of overde-velopment of tourism. As Mallorca attracts alarge number of domestic and foreigntourists, a noisy, dirty and over-commercial-ized atmosphere has become apparent. Thesecan also be good benchmarks for Turkey tocontrol overdevelopment in tourism in thenear future. Some problems in Turkey didnot appear in Mallorca (e.g. harassment,poor roads and driving, beaches, and air-con-ditioning) because Mallorca has improved itstourism infrastructure and services to reachstandards set in Europe.

Despite the strength of Turkey’s traditionof hospitality, a major complaint has alwaysbeen harassment by shopkeepers and restau-rateurs. A number of tourists complain that‘you get a lot of hassle when out shopping orlooking for somewhere to eat… Get sick ofhearing “yes please” constantly… If they leftyou alone, they would probably make moresales’. The open-ended comments frequently

included suggestions on how to solve theharassment problem. One respondent empha-sized that ‘… We would have bought muchmore if allowed to browse and not to bepestered continually!’. Similarly, another saidthat ‘… if they hassled a little less they wouldprobably sell more… I would have to thinkabout Marmaris because better nightlife andmore things to do… No hassle would be nice’.This provides a potential practical implicationto be investigated further. A comment aboutthis problem given by a customer may be sig-nificant: ‘It would be better if they just let youdecide if you wanted to go to their shops’.

Such evidence as over-commercialization,overcrowding, noise and dirty environmentcan be regarded as prime indicators of rapiddevelopments in Mallorcan tourism, a matureinternational tourism destination. As a conse-quence of the emergence of mass tourism, ithas attracted a high volume of tourists, butwith a low level of tourism income and spend-ing. To meet the increasing demand, accom-modation capacity has extended and widenedto other unspoilt areas causing environmen-tal deterioration. This also led to over-crowded beaches by the mid-1980s. BecauseMallorca has reached the mature stage of itsdestination life cycle, the local authority hasdecided to revitalize the image of Mallorca byestablishing a long-term planning policy.

The majority of complaints about overde-velopment in Mallorca are related to losingits original culture, nature and food. Onefirst-time tourist drew attention to the extentof this problem by underlining that ‘It is verycommercial… Losing its culture and her-itage… Like visiting a British resort in thesun…’. The next case addressed by anothertourist has links to support this statement: ‘…Food has not improved… Too much fast andeasy food is available… Nothing fancy or dif-ferent is available… Too many chips witheverything. Why do they think the English eatso many chips?’. One could speculate that thelocal cuisine has a limited number of dishesand it is, therefore, easy to find traditionalBritish food in Mallorca, e.g. fish and chips orBritish-style pubs. There are some customerswho observed the results of the implementa-tion of the recent tourism developmentplans: ‘… It has become more commercial-

298 M. Kozak


ized and is starting to detract from theisland’s original attraction. Happily, all newbuilding appears to be of low height, reduc-ing its detrimental effect’. Despite being per-ceived largely as a family holiday destination,Mallorca attracts a mixture of both young andmiddle-aged tourists in the summer seasonwhile older tourists prefer to go betweenOctober and May. This has raised a numberof potential problems particularly for thosewho travel as a family. Such resorts as Magalufand Arenal have become more popular withyounger people who tend to continue drink-ing and staying up late which disturbs others.

Although resorts in Turkey are muchyounger and not as over-commercialized asthose in Mallorca, there is already some evi-dence of a potential threat (dirtiness, crowds,noise and loss of culture and nature) in thefuture unless tourism development in Turkeyis controlled. Findings address the impor-tance of paying attention to improving thequality of resorts in Turkey before it reachesthe maximum development stage. Feedbackobtained from some repeat tourists is signifi-cant in identifying factors that lead to a desti-nation moving on to the next stage in thedestination life-cycle. Among these are a busyatmosphere and the lack of original cultureand nature. One points out that it is ‘…Getting too overcommercialized, spoiling nat-ural beauty as too many hotels are beingbuilt…’. The next one observes that ‘Turkishresorts are becoming too commercial and los-ing their cultural charm…’. The third obser-vation is very similar to one made aboutMallorca and signals the degree of potentialthreats for the tourism industry in Turkey inthe future: ‘It is busier with more tourists… Ithas become too English, e.g. prices inEnglish, restaurants named after English pro-grammes…’. If external benchmarking isbelieved to be worthy of consideration, this iswhat Turkey has to choose as a benchmarkand learn lessons from Mallorca.

In summary, such attributes as hospitality,plenty to do and see, level of prices, weatherconditions, and value for money were per-ceived to be among the most effective in thecompetitive position of Turkey. It is interest-ing to note that both hospitality and harass-ment were ranked as the first positive and

negative attributes, respectively. Despite thedifficulty of assessing it, this finding might beregarded as an indicator of rapid develop-ment in the country’s tourism industry.Those attributes positively affecting the com-petitiveness of Mallorca were hospitality,plenty to do and see, quality of beaches andthe sea, being family oriented, and suitabilityof accommodation facilities. AlthoughMallorca was perceived to be cleaner, it wasalso found to be dirtier than some other des-tinations. Therefore, it is not clear whetherthis attribute will be taken into considerationas a positive or negative element of theMallorcan tourism industry.

Competitor Analysis

A competitiveness set was defined by collect-ing data on the respondents’ visits to otherdestinations in the preceding 4 years (Table25.9). The objective was to ascertain whetherboth sample destinations were in the samecompetitive set. Except for the USA, the firstten destinations were in the Mediterraneanbasin, even though slight differencesappeared in the ranking lists. These destina-tions are the Canary Islands, Greece, France,Spain, Ibiza, Portugal, Cyprus, Italy, Turkey(or Mallorca) and Malta. The share of thesedestinations for those who visited Mallorcawas higher than for those who visited Turkey.The USA was ranked as the second most vis-ited destination by British respondents inter-viewed in Mallorca (25.2%), whereas it wasranked as the fourth by those interviewed inTurkey (18.0%). The share of other destina-tions such as long-haul (e.g. the Far East andAfrica) and short-haul (e.g. other Europeancountries) remained much smaller.

As a consequence, one could suggest thatthe study respondents had a much higher loy-alty to Mediterranean basin destinations andthe USA (particularly Florida) for the pur-pose of taking mainly ‘sea–sand–sun’ vaca-tions. Such an attempt at competitivenessanalysis could be significant for destinationcomparison research in order to have a betterunderstanding of competitors in the same setin a particular market and make decisionsabout whom and what to benchmark. For



instance, in the competitiveness sets discussedin this study, both Mallorca and Turkey wouldhave an opportunity to select their partner(s)from those in the Mediterranean basin.

Discussion

The starting point for this study is thepremise that, when tourists have visited multi-ple destinations, comparative surveys mightbe used to explore performance gaps. Thispremise is based on the assumption thattourists visiting multiple destinations wouldbe better able to identify gaps between theperformance of different destinations, e.g.what tourists perceive (destination attrib-utes), how they perceive (likes/dislikes), whattype of result is obtained for comparison (bet-ter/worse) and so on. The contribution ofthis assumption focuses on the significance ofchoosing a correct sampling method.Consumer behaviour research considers theexamination of the service strengths andweaknesses from the perspective of those whohave experienced them. If the purpose is tomeasure the performance of any destination,either host or partner, and use the term ‘per-formance ratings’, the sample must beselected from among those who have holidayexperiences at both destinations. Such directcomparison surveys can be used to directlycompare a destination’s performance withthat of another.

For this study, open-ended questionnaireswere used to compare self-reported, multipledestinations. This methodology obtainsdetailed information in tourists’ own wordsabout positive or negative holiday experiencesat visited destinations. This is also a usefulmethod for obtaining comments from touristsfor improvement. Some significant feedbackwas obtained from the results on the benefit ofusing open-ended questions in comparisonresearch and from the information that theresearcher was given by tourists in short face-to-face interviews at the study airports. Forexample, complaints about the noise of youngpeople after midnight reported by familygroups in Mallorca and the lack of air-condi-tioning systems reported by those in Turkeywere identified as areas needing improvement.

The use of open-ended questionnaires hasvaluable implications for research on thecomparative measurement of destination per-formance. This instrument was designed tomeasure one destination’s performance, notonly against a specific alternative destination,but also against other major competitors asidentified by the survey respondents. Thedestination authority is able to understand itsperceived performance against its main com-petitors. In addition, this instrument helps toidentify both positive and negative aspects ofthe destination’s tourism product. Forinstance, harassment was found to be themost serious problem in Turkey as comparedto other European destinations. Similarly,Turkey is perceived to have a better traditionof hospitality than several of its Europeancounterparts.

The proposed methodology is likely towork better if applied while the respondentsare still at the destination and in the pres-ence of interviewers, so as to prevent confu-sion and to yield fresh feedback eventhough a halo effect may occur. This is alsoa cost-effective way of conducting surveysamong tourists visiting international desti-nations. Moreover, the importance oftourists talking about holiday experiencesand the exchange of information with oth-ers after returning home should not beunderestimated. The destination has a greatopportunity to monitor its performancebecause it attracts a variety of tourist groupsrepresenting different geographical areasworldwide, many of whom may have also vis-ited other destinations.

However, this study also experiencedsome difficulties that need careful considera-tion in designing future research. First, it wasobserved that some respondents did not paysufficient attention when selecting the com-parison destination. For example, a fewrespondents tended to make a comparisonwith a destination they had visited 10–15years earlier, rather than within the 4-yearlimit. A direct quotation on the perceivedchanges in attitudes towards tourists sincethe latest visit better explains this: ‘… It isgood to see women out. This means thatEuropean women (like myself) get hassled abit less by men. It did seem dangerous to

300 M. Kozak


travel in Turkey in 1971’. It was furtherobserved that people are reluctant to make aclear decision on the question asking therespondents to identify the best destinationout of those they had visited in the preceding4 years. This is characterized by such typicalsentences as: ‘I cannot choose because theyare so different. Egypt is viewed as beingdirty and getting hassled but you visit thecountry for the famous pyramids and tombs.Turkey has less tourist attractions but overallthe beaches, facilities and nightlife are bet-ter’. Or the respondents like both: ‘Bothwere good in different ways. I would not sayone was better than the other’.

Finally, despite the fact that computertechnology has become very useful for con-tent analysis, some pitfalls still remain, e.g.concerns of objectivity, clarity and reliability.For example, the respondents were likely togive a general response to a question askingabout what they liked and disliked. Theytended to indicate a very small number ofattributes (reasons) for each question. Thereis no evidence to indicate whether thesewere the only attributes found to be betteror worse than in other destinations. The useof open-ended questionnaire is, therefore,

somewhat limited and needs furtherimprovement in future tourism research.

Conclusion

This study was designed to test the use of anopen-ended questionnaire to measure exter-nal performance of international tourist des-tinations. Briefly summarizing the results ofthe above analysis, the findings providegrounds for identifying areas where the studydestinations have strengths and weaknesses,not only against each other but also againstsome other competitor destinations. Thus,employing qualitative research methods hasseveral advantages in gaining a clear under-standing of tourists’ judgements in their ownwords and interpretation while examiningcomparative performance of tourist destina-tions. This type of assessment is useful partic-ularly for measuring the direction of touristperceptions (positive/negative or better/worse) and taking them into consideration inorder to evaluate tourists’ judgements aboutthe quality of facilities and services providedin one destination vis-à-vis that of other simi-lar destinations.


References

Bothe, C., Crompton, J.L. and Kim, S. (1999) Developing a revised competitive position for Sun/Lost City,South Africa. Journal of Travel Research 37, 341–352.

Calantone, R.J., di Benedetto, C.A., Halam, A. and Bojanic, D.C. (1989) Multiple multinational tourismpositioning using correspondence analysis. Journal of Travel Research 28, 25–32.

Crompton, J.L. (1992) Structure of vacation destination choice sets. Annals of Tourism Research 19, 420–434.Danaher, P.J. and Haddrell, V. (1996) A comparison of question scales for measuring customer satisfaction.

International Journal of Service Industry Management 7, 4–26.Decrop, A. (2001) The antecedents and consequences of vacationers’ dis/satisfaction: tales from the field.

In: Mazanec, J.A., Crouch, G.I., Brent Ritchie, J.R. and Woodside, A.G. (eds) Consumer Psychology ofTourism, Hospitality and Leisure, Vol. 2. CAB International, Wallingford, UK, pp. 333–348.

Driscoll, A., Lawson, R. and Niven, B. (1994) Measuring tourists’ destination perceptions. Annals of TourismResearch 21, 499–510.

Faulkner, B., Oppermann, M. and Fredline, E. (1999) Destination competitiveness: an exploratory exami-nation of South Australia’s core attractions. Journal of Vacation Marketing 5, 125–139.

Goodrich, J.N. (1977) Differences in perceived similarity of tourism regions: a spatial analysis. Journal ofTravel Research 16, 10–13.

Goodrich, J.N. (1978) The relationship between preferences for and perceptions of vacation destinations:application of a choice model. Journal of Travel Research 16, 8–13.

Haahti, A.J. (1986) Finland’s competitive position as a destination. Annals of Tourism Research 13, 11–35.Haahti, A. and Yavas, U. (1983) Tourists’ perceptions of Finland and selected European countries as travel

destinations. European Journal of Marketing 17, 34–42.Heath, E. and Wall, G. (1992) Marketing Tourism Destinations: a Strategic Planning Approach. John Wiley &

Sons, Canada.


Jackson, M.S., White, G.N. and Schmierer, C.L. (1996) Tourism experiences within an attributional frame-work. Annals of Tourism Research 23, 798–810.

Javalgi, R.G., Thomas, E.G. and Rao, S.R. (1992) US pleasure travellers’ perceptions of selected Europeandestinations. European Journal of Marketing 26, 45–64.

Johns, N. and Lee-Ross, D. (1995) Profile accumulation: a quality assessment technique for hospitalitySMEs. In: Teare, R. and Armistead, C. (eds) Services Management: New Directions and Perspectives. Cassell,London.

King, B. (1994) Australian attitudes to domestic and international resort holidays: a comparison of Fiji andQueensland. In: Seaton, A.V. et al. (eds) Tourism: the State of the Art. John Wiley & Sons, New York, pp.347–358.

King, R.L. and Woodside, A.G. (2001) Qualitative comparative analysis of travel and tourism purchase-con-sumption systems. In: Mazanec, J.A., Crouch, G.I., Brent Ritchie, J.R. and Woodside, A.G. (eds)Consumer Psychology of Tourism, Hospitality and Leisure, Vol. 2. CAB International, Wallingford, pp.87–105.

Kozak, M. (2003) Measuring comparative destination performance: a study in Spain and Turkey. Journal ofTourism and Travel Marketing, 13, 83–110.

Kozak, M. and Rimmington, M. (1999) Measuring destination competitiveness: conceptual considerationsand empirical findings. International Journal of Hospitality Management 18, 273–283.

Laws, E. (1995) Tourist Destination Management: Issues, Analysis and Policies. Routledge, New York.Mayo, E.J. and Jarvis, L.P. (1981) The Psychology of Leisure Travel: Effective Marketing and Selling of Travel

Services. CBI Publishing Company, Boston, Massachusetts.Melcher, A., Acar, W., DuMont, P. and Khouja, M. (1990) Standard-maintaining and continuous improve-

ment systems: experiences and comparisons. Interfaces 20, 24–40.Pearce, D.G. (1997) Competitive destination analysis in Southeast Asia. Journal of Travel Research 35, 16–24.Pearce, P.L. and Caltabiano, M.L. (1983) Inferring travel from traveller’s experiences. Journal of Travel

Research Fall, 16–20.Reilly, M.D. (1990) Free elicitation of descriptive-adjectives for tourism image assessment. Journal of Travel

Research 28, 21–26.Robson, C. (1993) Real World Research: a Resource for Social Scientists and Practitioner-Researchers. Blackwell,

Oxford.Zairi, M. (1996) Benchmarking for Best Practice: Continuous Learning through Sustainable Innovation.

Butterworth-Heinemann, Oxford.

302 M. Kozak


Chapter twenty-sixCross-cultural Behaviour Research In Tourism: a Case Studyon Destination Image

Metin Kozak,1 Enrique Bigné,2 Ana González3 and Luisa Andreu2

1School of Tourism and Hotel Management, Mugla University, 48000 Mugla, Turkey; 2Faculty of Business and Economy Studies, Department of Management and Marketing,University of Valencia, Avda. dels Tarongers s/n, 46022 Valencia, Spain; 3Faculty of Businessand Economy Studies, Department of Management and Marketing, University of León,Campus de Vegazana s/n, 24071 León, Spain

Abstract

Focused on cross-cultural research in tourism, this study aims to emphasize the significance of exploringcross-cultural differences in consumer behaviour and, in particular, destination image. In order to achievethis objective, first of all a theoretical framework is developed to introduce a research agenda on cross-cul-tural consumer behaviour in tourism. From a practical perspective, a case study is carried out based ontourism databases and two ad hoc studies conducted by the Valencian Tourist Board (Spain) at its tourisminformation offices. ANOVA and factor correspondence analysis are considered as the main techniques toanalyse the differences of destination image attributes taking into account the tourists’ country of resi-dence. Finally, conclusions and managerial implications are discussed.

Introduction

According to the World Tourism Organization(WTO), tourism has emerged as one of themost relevant sectors on a world level, as it isthe major source of wealth in a number ofcountries (WTO, 1997). Statistics provided bythe WTO emphasize the economic signifi-cance of tourism at the global level (WTO,2003). From an international perspective, themost popular destinations in the world in theyear 2000 were France, followed by Spain, the

USA, Italy and China. Looking at these figures,the need for cross-cultural research in coun-tries like Spain can be appreciated, whereinternational tourist arrivals have been experi-encing an increasing growth rate and, at thesame time, from a practical perspective, studiesdealing with cultural characteristics are scarce.

From an academic viewpoint, conductingcross-cultural studies in tourism has both itssupporters and its critics. On the one hand,proponents such as Pizam (1999) show thatthis type of research can be justified, as a great



deal of evidence suggests that nationalityinfluences tourist behaviour. Others, e.g. Plog(1990), have pointed to a dearth of researchrelated to the cultural differences and similar-ities of tourists, and have suggested that therapid globalization of the tourist phenome-non and its international nature warrants abetter understanding of the global tourist.Critics, on the other hand, such as Dann(1993) and Peabody (1985) highlight the limi-tations of using nationality and country of res-idence as segmentation variables in tourismresearch. They both suggest that tourism isnow well and truly a global phenomenon andgenerating and destination societies are nolonger culturally uniform.

Despite the criticisms, it is clear from a liter-ature review that cross-cultural research intourism is receiving increasing attention fromacademics. An examination of tourism litera-ture shows an increasing interest in cross-cul-tural research (Hudson and Ritchie, 2001).Specifically, one of the purposes for doingcross-cultural research is to explore other cul-tures, learn about them and to test cultural dif-ferences in tourism marketing contexts.Additionally, it is well known that, based uponproducts offered, one particular destinationmay attract customers from different nationali-ties. The investigation of potential cross-cul-tural differences and similarities betweenvarious consumer groups representing differ-ent cultures in tourism visiting a particular des-tination is important for destinationmanagement to learn the profile of its cus-tomers, their values, preferences and behav-iour, and to implement effective positioningand market segmentation strategies (Sussmannand Rashcovsky, 1997; Reisinger and Turner,1998; Pizam, 1999).

The identification of the image perceivedby the tourist is fundamental in determining adestination’s competitiveness (Britton, 1979;Mayo and Jarvis, 1981; Mathieson and Wall,1982; Ahmed, 1991; Andreu et al., 2000).Efforts to understand the factors drivingtourists to visit or that are influencing theirimage perceptions of a particular destinationand how likely it is to be different from thoseof others visiting other destinations could helpdestination management in setting marketingstrategies (Reisinger, 1997). Depending upon

the empirical findings, destination manage-ment would either promote attributes that bestmatch tourist motivations, or concentrate on adifferent market where tourist motivations anddestination resources match each other.

All these arguments justify the need forresearch in destination image from a cross-cul-tural view. Particularly, the objectives of thisresearch are as follows: (i) to review the con-cept of destination image as well as to high-light cross-cultural research studies in tourismbased on indirect and direct studies; (ii) toanalyse a case study based on internationaland national databases regarding Europeantravellers in a leading tourism destination inSpain and, in particular, in one specific formof tourism – sun and beach tourism – in whichSpain specializes. In order to achieve theseobjectives, first of all a theoretical framework isdeveloped to introduce a research agenda oncross-cultural consumer behaviour in tourism.Second, a case study is undertaken from anessential analysis, based on relevant data of theEuropean Travel Monitor, Frontur (IET, 2003)as well as ‘tourist-info surveys’ provided by theValencian Tourism Board (VTB).

Conceptual Background

Tourist destinations are accepted as being akey component of national or internationaltourism activities. Each destination offers avariety of products and services to attracttourists. However, each visitor also has theopportunity and freedom to choose from aset of destinations (Crompton, 1992).Research findings indicate that different fac-tors may have an influence on destinationchoice, or the attractiveness of one particulardestination (Mayo and Jarvis, 1981; Sirakayaet al., 1996). For instance, each visitor mayhave different motivations, preferences andimage perceptions for different destinations.Most studies of destination choice and touristbehaviour have been related to investigatingthe relationship between the image of desti-nation and tourist preferences for the placeas a tourist destination as well as dis/satisfac-tion with the destination. At this stage,Morrison (1989) presents two criteria, objec-tive and subjective, that help tourists to decide

304 M. Kozak et al.


which one meets their own criteria best. Theformer includes prices, locations, physicalcharacteristics of facilities or destinations andso on, whereas the image of the destination isconsidered as subjective criterion. Both objec-tive and subjective criteria are significantattributes in forming a destination image.

It is important to bear in mind that studiesof image and attitude are different concepts,despite the fact that both are largely used inthe field of marketing. Two people may havethe same images of a place, but may hold dif-ferent attitudes towards it, e.g. warm weather(Kotler et al., 1996). The place can be per-ceived to be warm (image), but one may notlike warm weather or travel to a place that iswarm (attitude). A number of image studieshave been carried out to explore positive andnegative perspectives of destinations on sev-eral attributes (Pearce, 1982; McLellan andFousher, 1983; Richardson and Crompton,1988; Embacher and Buttle, 1989; Echter andRitchie, 1991). Such research indicates thatdestination images influence tourist behav-iour (Hunt, 1975; Pearce, 1982).

Emphasizing the importance that imageshave upon the tourist, Hunt (1975) arguesthat the images, beliefs and perceptions thatpeople have about a destination can influ-ence the growth of a tourist area as much as,or even more than, tangible resources. Imagestudies play a key role in the marketing andpromotion of destinations, particularly forthose who have never been to the destinationbefore (Baloglu and McCleary, 1999).Therefore, research on consumer behaviourand destination marketing could possibly beconducted first of all in order to understandthe areas where the destination is suffering interms of its image; and methods can be devel-oped to construct a positive image and to sug-gest how to use this positive image to makepeople feel that the destination has its owndistinctive quality. Although it is claimed thatimage perceptions of destinations may notalways reflect the reality, unfortunately, theycould affect the destination choice of poten-tial tourists when deciding where to take avacation (Goodrich, 1978).

Previous studies have highlighted the vari-ations in the travel characteristics and behav-iour of tourists from different countries. In

an effort to classify the methods used whencarrying out cross-cultural research, Pizamand his colleagues categorize two types ofstudies: indirect and direct studies. The first,‘the indirect method’, refers to how ‘out-siders’, such as local residents, tour guides orentrepreneurs see tourists or, in other words,how they perceive differences in the behav-iour of tourists across various nationalities.The other, ‘the direct method’, aims atexploring whether any differences exist in thebehaviour, values or satisfaction levels oftourists representing different nationalities,and therefore reflects tourists’ opinionsabout themselves or their experiences. Ingeneral, researchers have previouslyemployed both methods; these are summa-rized in Tables 26.1 and 26.2.

On the one hand, a review of indirect stud-ies supports the proposition that national cul-tures have a moderating effect on touristbehaviour, although the research is based onsubjective perceptions. On the other hand,other research is developed by means of directmethods of cross-cultural comparison research.This type of research explores the similaritiesand differences between multiple groups inrelation to several vacation travel patterns,tourist satisfaction, tourist motivation andimage perceptions of the selected destinations.

Overall, direct studies have tended tofocus on information sources used by trav-ellers, destination choice, tourist expectationsand benefits received. The resulting datafrom all these studies reveal cultural differ-ences that provide theoretical support forexpanded research in the area of cross-cul-tural behaviour in tourism. Reviewing directstudies, there is a lack of empirical studiesregarding destination image. As the casestudy developed in this research focuses ondestination image, one of the purposes of thisresearch is to contribute in this issue.

Case Study: European Travellers to theComunidad Valenciana (Spain)

The case study is centred on the analysis oftourists in one particular region of Spain. Asmentioned before, Spain stands out as one ofthe major tourist receiving countries (51.7

Cross-cultural Behaviour Research in Tourism 305


million people in 2002), which allows thecountry to maintain its quota on a world level,in spite of emerging markets (IET, 2003;WTO, 2003). Among the principal countriessending tourists to Spain, the UK and German

markets stand out in first place (47.7%).Behind them, although at a considerable dis-tance, is France with 11.9% and then, withcomparatively lower figures, The Netherlands,Italy, Belgium and Portugal (IET, 2003).

306 M. Kozak et al.

Table 26.1. Indirect studies.

Countries Variables ofResearchers involved the study Sample Main contribution

Pizam and Jeong Korea, Japan and 20 behavioural 86 Korean Perceptions of (1996) USA characteristics of tour-guides Korean tour-guides

Japanese, American of tourists from and Korean tourists three nationalities

Pizam and Bulgaria, Hungary, A variety of travel Seven samples Opinions andTelisman-Kosuta Poland, Spain, UK, characteristics of residents in impressions of host(1996) USA, Yugoslavia each country communitiesPizam and Reichel Japanese, French, 20 behavioural 63 Dutch tour Cross-cultural(1996) Italian, US tourists characteristics guides tourist behaviour

in The Netherlands based on subjectiveperceptions

Reisinger and Australia and Japan National culture 250 Australian Cultural factorsTurner (1997) dimensions, holiday tourism influencing

experiences providers Japanese holidayexperiences inAustralia

Table 26.2. Direct studies.

Countries Variables of theResearchers involved study/datasets Sample Main contribution

Sussmann and French and Vacation travel 189 passengers Cross-cultural ofRashcovsky (1997) English Canadians patterns travelling by bus, leisure travel

train and plane dimensionsReisinger and Australia and Japan National culture 618 Asian tourists Cultural factorsTurner (1997) dimensions, visiting Australia influencing

holiday Japanese holidayexperiences experiences in

AustraliaYou et al. (2000) Japanese and UK Pleasure Travel 1200 (Japan) Cross-cultural study

long-haul pleasure Market Survey for and 1208 (UK) on push and pulltravellers Japan (1995) and factors

for UK (1996)Crotts and Japan, UK, In-flight surveys of 80,000 tourists National culture onErdmann (2000) Germany, France, overseas tourists to the USA consumer

Brazil and Taiwan to the USA (1996, evaluations1997, 1998)

Hudson and Canada, USA and Environmentally 111 Canadians, Identification ofRitchie (2001) UK skiers friendly ski 116 Americans, customers who are

destination, 105 British skiers WTP more for anWTP, socio- E-F productdemographics


In respect to destinations preferred by for-eign tourists, traditionally the Balearic and theCanary Islands stand out as being the mostimportant, and also the Mediterranean coast-line: Catalonia, Andalusia, and the ValenciaCommunity (known in Spain as theComunidad Valenciana). The last two haveshown particularly important increases inrecent years (IET, 2003), a fact that ratifies theleadership that continues to be maintained bySpain in the typology of sun and beaches, inwhich the country specializes (www.comunidadvalenciana.com). In respect to thedestinations just mentioned, the ComunidadValenciana deserves particular emphasis here,having received nearly 5 million foreigntourists (approximately 10% of the total num-ber of tourists to Spain). Generally speaking,the principal outbound markets that come tothis community are approximately representa-tive of those visiting all of Spain, taken as awhole. Having justified the importance of con-ducting cross-cultural research in this area, thespecific objectives of the case study aretwofold: one is to analyse the profile of foreigntourists who visit the Comunidad Valenciana,and the other would be to analyse the per-ceived image that tourists have of this region,taking into consideration a representative sam-pling of the different nationalities that visit thetourist destination under investigation.

In order to achieve the above-mentionedobjectives, the case study is carried out basedon tourism databases and ad hoc studies con-ducted by the Valencian Tourist Board (VTB).One of the secondary sources of informationused is the European Travel Monitor (ETM)database, the objective of which is to providecomparable data on European travel behav-iour. The ETM is a continuous survey measur-ing all overnight trips made by the adultinhabitants of up to 33 European countries –irrespective of the reason for the trip. For thispurpose, Europeans are representatively inter-viewed about trips they have undertaken dur-ing the last few months in more than 150 wavesper year. This totals to approximately 400,000interviews per year. The results obtained fromthese polls representative of the population areextrapolated for the total population (aged 15years and above) of each country. Designed asa basic tool for marketing decisions in tourism,

the objective of the European Travel Monitoris to provide comparable data on Europeantravel behaviour, as well as on the develop-ment of travel patterns on a pan-Europeanbasic. Therefore, the ETM has been continu-ously recording all trips abroad taken byEuropeans with at least one overnight stay,including holiday trips, business trips, visits tofriends and/or relatives and other privatetrips. Interviews are carried out by telephoneusually from a central location using a CATIsystem (computer-aided telephone interview-ing). All data relate to trips, not to people. Ifone person makes three trips, these arecounted three times in the total volume.

Profiles of European travellers

Concerning the first objective, the profile oftourists to the Comunidad Valenciana isshown in Table 26.3. Specifically, based on theETM, data gathered for five key tourism mar-kets is shown (Germany, The Netherlands,Belgium, France and the UK) regarding theirvisits to the Comunidad Valenciana and thefollowing variables: purpose of trip, type ofholiday, accommodation, mode of transport,organization, travel spending, length of stay,travel season, degree of holiday satisfactionand intention to repeat. As far as trip purposeis concerned, it stands out that the majority ofthe tourists come for holidays, and mainly thetype of holiday is sun and beach, and most ofthem take trips lasting 4 or more nights. Ingeneral they are very satisfied with their holi-days and the intention to repeat is high.

Apart from the ETM market research, thedata for the case study was gathered from twomain and complementary sources. First ofall, the statistical study entitled ‘over-the-bor-der tourist movements’ (FRONTUR) carriedout by the Spanish Institute for TourismStudies based on interviews conducted atfrontiers (IET, 2003). Second, the 1999 and2000 ‘Survey of Visitors to the ComunidadValenciana’ was undertaken. Annually, theVTB surveys more than 1000 overseas touristsin Tourism Information Offices (Tourist-Info) during the summer, with self-adminis-tered questionnaires consisting of thefollowing variables: introductory variables



(i.e. first time of the visit); variables relatedto the tourist behaviour (i.e. mode of orga-nizing the holiday, through travel agent, ontheir own, etc.), accommodation used, andmeans of transport used; future behaviour

intentions (i.e. intention to repeat); socio-demographic variables (age, gender, place ofresidence, educational level) and the per-ceived destination image based on 19 attrib-utes (parking space, promenades, rest areas,

308 M. Kozak et al.

Table 26.3. European travel to the Comunidad Valenciana: a comparison of key outbound countries.

Variables Belgian British French German Netherlands

Volume of tripsa 288,000 1,300,000 613,000 781,000 238,000Trip purpose

Holidays 100% 69% 61% 72% 98%Business – 13% 24% – 1%Other – 18% 15% 28% 1%

Type of holidaySun and beach 92% 60% 66% 100% 66%Touring 8% 9% 7% – 5%City holiday – 15% – – –Private occasion – – 27% – –Sporting – – – – 9%Others – 16% – – 20%

AccommodationHotel 16% 47% 36% 5% 15%Other paid for 58% 35% 53% 61% 54%Private 18% 14% 11% 34% 32%Others 7% 4% – – –

Mode of transportPlane 42% 85% 35% 35% 52%Car 58% 6% 61% 42% 22%Others – 9% 4% 23% 26%

OrganizationPackage 11% 39% 5% – 38%Other pre-booking 64% 60% 63% 55% 42%No pre-booking 25% 1% 32% 45% 20%

Travel spendingb €787 €1,124 €1,009 €1,153 €905Length of stayc

Short trips 0% 18% 34% 7% 10%Longer trips 100% 82% 66% 93% 90%

Travel seasonSummer 51% 68% 50% 95% 58%Winter 49% 32% 50% 5% 42%

Holiday satisfactionYes, very nice 90% NA 100% 100% 96%So so 10% NA – – 4%No, not really 0% NA – – 1%

Intention to repeatYes 79% NA 94% 53% 66%Maybe 13% NA 6% 12% 31%Rather not 8% NA 0% 35% 2%

aTrips to Valencia with at least one overnight stay.bAverage expenditure per trip (and person). These costs included all expenses, means transport,accommodation, food and other expenses.cShort trips (1–3 nights) and longer trips (4+ nights).Source: European Travel Monitor (2000).


green zones, recreational and sports areas,accommodation, restaurants, bars/cafés,shops/stores, cultural and leisure offering,safety, traffic flow, tourist signs, cleanlinessand conservation, rubbish removal, beachcleanliness and conservation, state of theroads, ease of access to primary tourist sites,and absence of noise). A five-point scaleranging from ‘very low’ (1) to ‘very high’ (5)was used to evaluate how tourists perceiveeach destination attribute.

Based on the research conducted by theVTB in tourist-information offices during1999 and 2000, the samples of both studiesare: (i) with regard to the year 1999, the totalsample is 2879. Of these, 556 are Spanish, 469British, 322 German, 814 French and 633from the rest of the world. (ii) With referenceto the year 2000, the total sample is 2511. Ofthese, 510 are Spanish, 421 are British, 336are German, 664 are French and 580 fromthe rest of the world. The distribution of thesampling in reference to the socio-demo-graphic variables is very similar for the yearsanalysed, as well as in relation to the differentcountries-of-residence of tourists visiting theComunidad Valenciana. Specifically, the vari-able gender is distributed quite equally in thesampling analysed, but with a slightly higherpercentage of males over females in all cases,except for the French, in the year 1999.Regarding age, the most important volume ofvisitors is concentrated in the interval 25–44years, followed very closely by tourists in the45–65 years age group. The level of studiesalso shows a distribution by age, as well as bycountry, which is very homogeneous: pre-dominantly university studies, followed by sec-ondary studies. Finally, although this is not asocio-demographic variable, it is important toknow whether this was the first time the trav-ellers had visited the Comunidad Valenciana,given that this aspect could possibly deter-mine in the perception of the destination. Inthis case, for the year 1999, higher percent-ages were observed in tourists from the UK,Germany and the rest of the world, who werevisiting the Comunidad for the first time,when compared to Spanish and Frenchtourists. For the year 2000, this figure wasonly higher than the percentage of Spanishtourists, who had already visited the area,

when compared to the rest of the tourists,irregardless of the country-of-residence. Evenso, in all cases the percentages were in thearea of 50%.

The authors recognize that there are cer-tain limitations to the methodology whichmust be stated. From a theoretical point ofview, the analysis and measurement of touristdestination image through profiling destina-tion attribute is the approach that is most fre-quently applied. However, an importantlimitation of this approach is the heterogene-ity of the attributes used, rendering compar-isons between the different destinationsdifficult (Ruiz et al., 1999). In terms of thisspecific study, a pilot study would have allowedfor the identification of those areas that couldhave been investigated in greater depth.

Tourist destination image: cross-culturalstudy

With reference to the second objective of thecase study, this research focuses on the differ-ences among tourists regarding the image ofthe Comunidad Valenciana. As can be con-cluded based on the research overview, fewcross-cultural studies have been carried outon how people from varying cultural back-grounds differ in their perceived destinationimage. The present case study is based on theanalysis of destination image that differenttourists perceived of the ComunidadValenciana. Specifically, the differences ofdestination image depending on the tourists’country-of-residence (Spain, UK, France,Germany, and the rest of the world) areanalysed. Therefore, a direct method is used.With regard to the analysis of destinationimage from a cross-cultural point of view, itwas conducted using multivariate techniques:analysis of variance (ANOVA) for each of thedestination attributes, and factor correspon-dence analysis.

ANOVA of the destination attributes

First of all, an ANOVA for each of the desti-nation attributes of the study was carried outin order to investigate the influence of coun-try-of-residence on the perceived destination



image. As shown in Tables 26.4 and 26.5, forthe years 1999 and 2000 respectively, signifi-cant differences are highlighted (F-value).Furthermore, by means of a post hoc analy-sis, it was possible to determine whether thecountry-of-residence affected the evaluationof destination attributes. Regarding bothVTB databases, it became evident that thecountry-of-residence is related to the percep-tion of the Comunidad Valenciana as atourist destination. In other words, the per-ceived image of this tourist destination is nothomogeneous; rather, it is possible to iden-tify differences on the basis of the country.For instance, based on the 1999 database(Table 26.4), French tourists who visited theComunidad Valenciana (FR, n = 814) evalu-ated significantly lower (P �0.05) the follow-

ing attributes in comparison to Britishtourists (UK, n = 469): parking space, rest,cleanliness and conservation, rubbishremoval, beach cleanliness and conservation,and absence of noise.

Regarding the 2000 database (Table26.5), Spanish tourists who visit theComunidad Valenciana (SP, n = 510) evalu-ated most of the attributes lower than theother segments. Specifically, differences werefound in comparison to British (UK),German (GE) and the rest of the world (RW)regarding parking space (SP�GE, SP�RW,SP�UK), recreational and sports (SP�GE,SP�UK, SP�RW), accommodation (SP�UK,SP�GE, SP�RW), as well as restaurants,bars/cafés and shops/stores (SP�RW,SP�GE, SP�UK).

310 M. Kozak et al.

Table 26.4. Analysis of variance (ANOVA) by country-of-residence (year 1999).

SP, UK, FR, GE, RW,Destination attributes n = 556 n = 469 n = 814 n = 322 n = 718 F a Post hoc testb

Parking space 3.49 3.89 3.58 3.64 3.75 8.16 ** SP�RW, SP�UK, FR�UKPromenades 4.18 4.33 4.23 3.95 4.29 9.54 ** GE�SP, GE�FR, GE�RW,

GE�UKRest 3.96 4.05 3.76 3.97 4.07 9.25 ** FR�SP, FR�UK, FR�RWGreen zones 3.65 3.94 3.77 3.67 3.92 7.28 ** SP�RW, SP�UK,

GE�RW, GE�UKRecreational and 3.93 4.01 3.99 3.96 4.14 3.79 ** SP�RWsportsAccommodation 4.34 4.53 4.48 4.45 4.44 3.84 ** SP�FR, SP�UKRestaurants 4.37 4.56 4.48 4.49 4.48 4.56 ** SP�UKBars/cafés 4.42 4.61 4.56 4.53 4.56 6.01 ** SP�FR, SP�RW, SP�UKShops/stores 4.24 4.30 4.33 4.29 4.38 2.25Cultural and leisure 4.10 4.17 4.20 4.10 4.22 1.81line-upSafety 4.33 4.39 4.43 4.38 4.47 2.53 * Non-SDTraffic flow 3.62 3.77 3.65 3.54 3.69 2.06Cleanliness and 4.00 4.14 3.84 3.84 4.12 11.98 ** FR�RW, FR�UK, conservation GE�RW, GE�UKRubbish removal 4.15 4.24 4.05 4.13 4.22 3.97 ** FR�RW, FR�UKTourist signs 4.14 4.15 4.00 3.98 4.12 3.40 ** Non-SDBeach cleanliness 4.29 4.47 4.31 4.37 4.44 4.95 ** SP�RW, SP�UK, FR�UKand conservationState of the roads 4.01 4.10 4.24 4.13 4.29 9.67 ** SP�FR, SP�RW, UK�RWEase of access to 4.19 4.25 4.18 4.21 4.33 3.75 ** FR�RWprimary tourist sitesAbsence of noise 3.47 3.72 3.44 3.71 3.68 7.16 ** FR�RW, FR�GE, FR�UK,

SP�UK

a* P �0.05; **�0.01.bSignificant differences (SD) in the post hoc test (Scheffé): SP (Spain), UK (United Kingdom), FR(France), GE (Germany), RW (rest of the world).


Factor correspondence analysis

Secondly, a factor correspondence analysiswas carried out on the data for the 19 destina-tion attributes and five countries-of-residence,using the Data Theory Scaling System Group(DTSS), available with SPSS 11.0 for windows.Findings are presented separately based onthe 1999 and 2000 VTB databases, asexplained below.

In reference to the 1999 VTB database, twofactors emerged that were able to account forthe majority of the variance (71%) – the firstfactor accounts for 39% of the variance, andthe second for 32%. In order to understandthe dimensions, it is important to analyse towhat extent the different points (attributesand countries) contribute to the inertia ofeach dimension. Firstly, with regard to attrib-


Table 26.5. Analysis of variance (ANOVA) by country-of-residence (year 2000).

SP, UK, FR, GE, RW,Destination attributes n = 510 n = 421 n = 664 n = 336 n = 580 Fa Post hoc testb

Parking space 3.27 3.80 3.49 3.61 3.79 15.26 ** SP�GE, SP�RW, SP�UK,FR�RW, FR�UK

Pedestrian roads 4.17 4.25 4.19 3.97 4.36 9.91 ** GE�FR, GE�UK,GE�RW, SP�RW,FR�RW

Rest 3.93 3.99 3.88 3.96 4.24 10.93 ** FR�RW, SP�RW,GE�RW, UK�RW

Green zones 3.57 3.71 3.77 3.68 4.00 10.11 ** SP�RW, GE�RW,UK�RW, FR�RW

Recreational and sports 3.73 4.03 3.93 3.97 4.06 7.82 ** SP�FR, SP�GE, SP�UK,SP�RW

Accommodation 4.21 4.41 4.38 4.43 4.45 6.36 ** SP�FR, SP�UK, SP�GE,SP�RW

Restaurants 4.31 4.58 4.47 4.51 4.47 7.32 ** SP�RW, SP�FR, SP�GE,SP�UK

Bars/cafés 4.35 4.62 4.47 4.53 4.52 8.41 ** SP�RW, SP�GE, SP�UK,FR�UK

Shops/stores 4.20 4.45 4.29 4.42 4.39 7.94 ** SP�RW, SP�GE, SP�UK,FR�UK

Cultural and leisure 4.02 4.29 4.18 4.20 4.21 5.03 ** SP�RW, SP�UKline-upSafety 4.25 4.20 4.24 4.19 4.25 0.37Traffic flow 3.47 3.75 3.60 3.58 3.85 9.36 ** SP�UK, SP�RW,

GE�RW, FR�RWCleanliness and 3.89 4.05 3.91 4.01 4.09 4.25 ** SP�RW, FR�RWconservationRubbish removal 3.98 4.09 3.99 3.97 4.15 2.64 * Non-SDTourist signs 4.01 4.12 3.93 3.89 4.19 6.88 ** GE�RW, FR�RWBeach cleanliness 4.26 4.39 4.24 4.20 4.39 4.25 ** Non-SDand conservationState of the roads 3.91 4.07 4.33 4.02 4.23 17.75 ** SP�RW, SP�FR,

GE�RW, GE�FR,UK�RW

Ease of access to 4.07 4.32 4.23 4.14 4.32 7.64 ** SP�RW, SP�UKprimary tourist sitesAbsence of noise 3.42 3.70 3.52 3.59 3.76 6.40 ** SP�UK, SP�RW, FR�RW

a* P �0.05; **�0.01.bSignificant differences (SD) in the post hoc test (Scheffé): SP (Spain), UK (United Kingdom), FR(France), GE (Germany), RW (rest of the world).


utes, it is noticed that, on the one hand, rest(0.14) and cleanliness and conservation(0.16), and in the opposite area, the state ofthe roads (0.18) contribute to the inertia(explained variance) of the dimension 1.Secondly, promenades (0.35) and absence ofnoise (0.28) in the reverse side contribute tothe inertia of the second dimension.Regarding the countries, the UK (0.35) andFrance (0.61) on the opposite side, contributeto dimension 1. Concerning the seconddimension, Spain (0.17) and Germany (0.75)contribute to the inertia of this dimension.

With reference to the 2000 VTB database,two factors emerged that were able toaccount for the majority of the variance(70%) – the first factor accounts for 40% ofthe variance, and the second for 30%.Analogously, in order to understand thedimensions, we analysed to what extent thedifferent points (attributes and countries)contribute to the inertia of each dimension.Firstly, with regard to attributes, we were ableto note that parking space (0.38), and in theopposite area, both promenades (0.15) andsafety (0.15) contribute rather high to the

inertia (explained variance) of dimension 1.Secondly, green zones (0.11) and restaurants(0.12) – in the reverse side – contribute tothe inertia of dimension 2. In relation to thecountries, Spain (0.58), as well as the UK(0.16) and Germany (0.20) on the oppositeside, contribute to dimension 1. Regardingthe second dimension, Germany (0.23) and,on the reversal, the rest of the world (0.68)contribute to the inertia of this dimension.

Correspondence analysis results can beportrayed in perceptual maps. Going furtherin the analysis, to interpret their axes, it isadvisable to analyse the contribution (CTR)of dimensions to the inertia of each row(attributes) and column (countries) points(Bigné and Vila, 1999). Therefore, weanalysed which points (attributes and coun-tries) are better explained by the first and sec-ond dimension. With this objective, thehigher values left–right for dimension 1 (hor-izontal axis), and the higher values top–bot-tom for dimension 2 (vertical axis) wereselected. These figures are shown in Tables26.6 and 26.7, and represent the percentageof variation of attribute and country pointsexplained for each dimension.

312 M. Kozak et al.

Table 26.6. Contribution of dimensions to the inertia of each attribute point.

1999 2000

Destination attributes Dimension 1 Dimension 2 Dimension 1 Dimension 2

Parking space 0.35 0.08 0.85 0.13Promenades 0.00 0.95 0.62 0.24Rest 0.61 0.08 0.07 0.45Green zones 0.07 0.13 0.00 0.45Recreational and sports 0.27 0.00 0.86 0.03Accommodation 0.34 0.14 0.09 0.59Restaurants 0.20 0.32 0.09 0.89Bars/cafés 0.46 0.18 0.11 0.79Shops/stores 0.84 0.03 0.22 0.59Cultural and leisure line-up 0.90 0.07 0.17 0.77Safety 0.88 0.02 0.70 0.26Traffic flow 0.02 0.70 0.21 0.76Cleanliness and conservation 0.66 0.29 0.01 0.09Rubbish removal 0.45 0.01 0.74 0.02Tourist signs 0.11 0.33 0.35 0.29Beach cleanliness and conservation 0.18 0.75 0.54 0.00State of the roads 0.75 0.02 0.05 0.01Ease of access to primary tourist sites 0.06 0.05 0.01 0.00Absence of noise 0.27 0.72 0.65 0.29


Looking at Table 26.6, for the year 1999,the first dimension, or horizontal axis,explains in the negative side attributes suchas parking space (CTR = 35%), rest (CTR =61%), cleanliness and conservation (CTR =66%), as well as rubbish removal (CTR =45%). On the positive side, it explains attrib-utes such as recreational and sports (CTR =27%), accommodation (CTR = 34%), restau-rants (CTR = 20%), bars/cafés (CTR = 46%),shops/stores (CTR = 84%), cultural andleisure line-up (CTR = 90%), safety (CTR =88%), and state of the roads (CTR = 75%).The second dimension, or vertical axis,explains in its negative side attributes such asrestaurants (CTR = 32%), beach cleanlinessand conservation (CTR = 75%) and absenceof noise (CTR = 72%). On the positive area, itexplains attributes such as promenades (CTR= 95%), traffic flow (CTR = 70%), cleanlinessand conservation (CTR = 29%) and touristsigns (CTR = 33%).

The first dimension, or horizontal axis,explains in the negative side countries (seeTable 26.7) such as the UK (CTR = 63%) and,on the positive part, France (CTR = 85%).The second dimension, or vertical axis,explains in its negative side, Germany (CTR =96%) and on the positive area, Spain (CTR =31%). The 1999 perceptual map obtainedbased on the two dimensions is shown in Fig.26.1. The attribute-based perceptual mapshows the relative proximities of both coun-tries and attributes in a joint space.

Regarding the year 2000 (Table 26.6), thefirst dimension, or horizontal axis, explains inthe negative side attributes such as parkingspace (CTR = 85%), recreational and sports(CTR = 86%), as well as absence of noise(CTR = 65%). On the positive side, it explains

attributes such as promenades (CTR = 62%),safety (CTR = 70%), rubbish removal (CTR =74%), beach cleanliness and conservation(CTR = 54%) and tourist signs (CTR = 35%).The second dimension, or vertical axis,explains, in its negative side, attributes such asrest (CTR = 45%), green zones (CTR = 45%),traffic flow (CTR = 75%) and tourist signs(CTR = 29%). On the positive area, it explainsaccommodation (CTR = 59%), restaurants(CTR = 89%), bars/cafés (CTR = 79%),shops/stores (CTR = 59%), cultural andleisure line-up (CTR = 77%).

The first dimension, or horizontal axis,explains in the negative side countries (seeTable 26.7) such as the UK (CTR = 46%) andGermany (CTR = 44%). On the positive side,it explains countries such as Spain (CTR =82%). The second dimension, or vertical axis,explains in its negative side, the rest of theworld (CTR = 93%) and on the positive area,it explains countries such as Germany (CTR =38%). The 2000 perceptual map obtainedbased on the above-mentioned dimensions isportrayed in Fig. 26.2.

After having obtained the perceptual map,it is possible to interpret the correspondenceanalysis results, which are consistent with pre-vious ANOVA findings. First of all, it is possi-ble to analyse the relative similarity ordissimilarity of countries and the associatedattributes. Low distance indicates high simi-larity, and high distance has an oppositeinterpretation.

Regarding the year 1999 (Fig. 26.1), thefirst dimension on the horizontal axis com-pares the destinations of the UK and France.With respect to the attributes, parking space,cleanliness and conservation, rubbishremoval, beach cleanliness and conservation,


Table 26.7. Contribution of dimensions to the inertia of each country point.

1999 2000

Destination attributes Dimension 1 Dimension 2 Dimension 1 Dimension 2

Spain 0.04 0.31 0.82 0.01UK 0.63 0.02 0.46 0.00France 0.85 0.07 0.08 0.11Germany 0.01 0.96 0.44 0.38Rest of the world 0.07 0.03 0.01 0.93


314 M. Kozak et al.

SP

RWUK

Cleanliness andconservation

Promenades

Traffic flowTourist signs

Greenzones

0.30

0.20

0.10

0.000.00�0.05�0.15�0.20 �0.10 0.05 0.10 0.15 0.20 0.25

Dim

ensi

on

2�0.10

�0.20

�0.30

Rubbishremoval

Parking spaceRestBeach cleanlinessand conservation

Absence of noise

GE

Restaurants

Ease ofaccess

AccommodationBars/Cafés

Shops/Stores SafetyState of the roads

Recreationaland sports

Cultural andleisure line-up

FR

Dimension 1

Fig. 26.1. Perceptual mapping with correspondence analysis (year 1999).

GE

Restaurants

Accommodation

0.20

0.15

0.10

0.00–0.10–0.30–0.40 –0.20 0.00 0.10 0.20 0.30

Dim

ensi

on

2

Absence of noise

Parking space

State of the roads SP

Dimension 1

0.05

–0.05

–0.10

–0.15

–0.20

Shops/Stores Bars/Cafés

FR

Cultural and leisure line-up

Recreationaland sports

Cleanliness andconservation Ease of

accessBeach cleanlinessand conservation

Safety

Rubbish removal

Tourist signs

PromenadesRest

Green zones

UK

Traffic flow

RW

Fig. 26.2. Perceptual mapping with correspondence analysis (year 2000).


and absence of noise, these are situated farfrom the position of France, compared to theUK position where these attributes stand out.This finding can be also appreciated in themean values and significant differences thatwere obtained in the ANOVA (Table 26.4).Concerning these attributes, the relativeproximity of those attributes to the UK sup-port the higher evaluation of the UK samplein comparison to the French sample. Withreference to dimension 2, Germany andSpain show differences in the evaluation ofpromenades. This destination attribute has alow distance with Spain and, in contrast, ahigh distance with Germany. From theANOVA, the significant difference betweenboth countries regarding promenades(GE�SP, P �0.01) is highlighted.

In the 2000 perceptual mapping, as can beseen from Fig. 26.2, the first dimension onthe horizontal axis compares the destinationsof Spain with Germany and the UK. Withrespect to the attributes, it is important tomention the clear proximity with which fourattributes (restaurants, shops/stores,bars/cafés, accommodation) are perceived,indicating their similarity in the image as per-ceived by the tourists. Specifically, Germansand British perceived those attributes asbeing higher, in comparison with the Spanishsample. Based on the ANOVA post hoc analy-sis, UK and Germany show significant differ-ences with Spain (SP�UK, SP�GE).Regarding the second dimension, the highdistance between the rest of the world andGermany is noted. In respect to the attrib-utes, the perceptual maps show the relativesimilarity of the rest of the world and the fol-lowing attributes: rest, green zones, trafficflow and tourist signs. Consistently, in thepost hoc analysis, a higher perception of therest of the world sample (GE�RW) regardingrest, green zones, traffic flow and tourist signsis corroborated.

Conclusions

This chapter presents a case study about desti-nation image from a cross-cultural perspective.Initially, as far as the first objective is con-cerned, and based on secondary sources such

as the ETM, the profile of foreign tourists whovisit the Comunidad Valenciana is mainly forsun and beach holidays. Holiday satisfaction isvery high and the intention to repeat,although also quite high, varies according tonationalities. Next, as far as the second objec-tive referring to the perceived image, the fol-lowing principal findings can be mentioned.

The perception of tourists regarding a spe-cific tourist destination (i.e. ComunidadValenciana) is not homogeneous. Based onANOVA and perceptual mapping with corre-spondence analysis, this research reveals sig-nificant differences in the destination imageattributes. Figures 26.1 and 26.2 synthesizethe position of different countries, which canbe interpreted according to the relative simi-larity or dissimilarity of countries and theassociated attributes. From a managerial per-spective, the potential of exploring thetourism destination image taking into consid-eration the different generating markets is animportant issue that needs to be dealt with.

Despite their relative simplicity, perceptualmaps are powerful strategic tools in that theyallow managers to absorb a tremendousamount of data in a visual format (Luckett etal., 1999). Although perceptual mapping is anextremely powerful heuristic, it does have anumber of limitations. Perceptual maps arenot dynamic. They represent a static view of acompetitive marketplace at a particular pointin time and do not provide data regarding howthese countries or brands achieve their currentpositions (Luckett et al., 1999). Another limita-tion is that the correspondence analysis solu-tion is conditioned on the set of attributesincluded (Hair et al., 1995). It assumes that allattributes are appropriate for how consumersfrom different countries evaluated destinationimage, and that the same dimensionalityapplies to each country-of-residence. In short,the advantages of the joint plot of attributesand countries-of-residence must always beweighed against the inherent interdependen-cies that exist and the potentially biasingeffects of a single inappropriate attribute orcountry. However, the method provides a pow-erful tool for gaining managerial insight intocross-cultural destination image, i.e. the rela-tive position of countries and the attributesassociated with those positions.



This study has gone some way towardsidentifying the position of different touristson the basis of their country-of-residence,regarding their perceived image of a destina-tion. Because this study was limited to a par-ticular destination that mainly receivesEuropean travellers, it is not possible to gen-eralize its findings to other worldwide tourismdestinations. Therefore, the authors proposeto extend further research in new destina-tions and generating tourism markets outsidethe European frontiers in order to advancethe understanding of cross-cultural touristbehaviour.

Apart from these practical findings, theauthors would like to point out the chal-lenges of following this line of research.Tourist destinations attract tourists from dif-ferent cultures and countries; it is not reason-able to take into consideration only onespecific group of customers. A comparativeanalysis between groups is required in order

to better demonstrate the importance of anunderstanding of the variables underlyingthese national differences. There is one ques-tion that still remains here: does this appearto exist as a result of cultural differences ornational differences? Or does each countryor nation represent a unique culture that dis-tinguishes it from others? Do cultures need tobe identified with nations? Or can various cul-tures be distinguished within each nation?Could there even be similarities betweenregions of different nationalities? This area oftourism research is very new and requiresmuch attention for exploration in the future.

Acknowledgements

The authors would like to thank theValencian Tourism Board (AgenciaValenciana del Turismo) for the collabora-tion that they have provided.

316 M. Kozak et al.

References

Ahmed, Z.U. (1991) The influence of the components of a state’s tourist image on product positioningstrategy. Tourism Management 12, 331–340.

Andreu, L., Bigné, E. and Cooper, C. (2000) Projected and perceived image of Spain as a tourist destina-tion for British travellers. Journal of Travel and Tourism Marketing 9, 47–67.

Baloglu, S. and McCleary, K.W. (1999) U.S. international pleasure travelers’ images of four Mediterraneandestinations: a comparison of visitors and nonvisitors. Journal of Travel Research 38, 144–152.

Bigné, E. and Vila, N. (1999) Aplicación del análisis multidimensional y del análisis factorial de correspon-dencias. In: Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C. (eds) Análisis multivariante, 5thedn. Prentice Hall Iberia, Madrid, pp. 595–608.

Britton, R.A. (1979) The image of the third world in tourism marketing. Annals of Tourism Research 6,318–329.

Crompton, J.L. (1992) Structure of vacation destination choice sets. Annals of Tourism Research, 19,420–434.

Crotts, J.C. and Erdmann, R. (2000) Does national culture influence consumers’ evaluation of travel ser-vices? A test of Hofstede’s model of cross-cultural differences. Managing Service Quality 10, 410–419.

Dann, G. (1993) Limitations in the use of nationality and country of residence variables. In: Pearce, D.G.and Butler, R.W. (eds) Tourism Critiques and Challenges. Routledge, London, pp. 88–112.

Echtner, C.M. and Brent Ritchie, J.R. (1991) The meaning and measurement of destination image. TheJournal of Tourism Studies 2, 2–12.

Embacher, J. and Buttle, F. (1989) A repertory grid analysis of Austria’s image as a summer vacation desti-nation. Journal of Travel Research 27, 3–7.

European Travel Monitor (2000) European Travel Monitor Travel to Valencia in 2000. IPK International,Munich.

Goodrich, J.N. (1978) The relationship between preferences for and perceptions of vacation destinations:application of a choice model. Journal of Travel Research, 16, 8–13.

Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C. (1995) Multivariate Data Analysis with Readings, 5thedn. Prentice Hall International, Englewood Cliffs, New Jersey.

Hunt, J.D. (1975) Image as a factor in tourism development. Journal of Travel Research 13, 1–7.


Hudson, S. and Ritchie, J.R.B. (2001) Cross-cultural tourist behavior: an analysis of tourist attitudestowards the environment. Journal of Travel and Tourism Marketing 10, 1–22.

IET (2003) El turismo en España durante 2002. Instituto de Estudios Turísticos, Madrid.Kotler, P., Bowen, J. and Makens, J. (1996) Marketing for Hospitality and Tourism. Prentice Hall, Englewood

Cliffs, New Jersey.Luckett, M., Ganesh, J. and Gillett, P. (1999) Quantitative tools in tourism research: an application of per-

ceptual maps. In: Pizam, A. and Mansfeld, Y. (eds) Consumer Behavior in Travel and Tourism. TheHaworth Press, New York, pp. 307–333.

Mathieson, A. and Wall, G. (1982) Tourism: Economic, Physical and Social Impacts. Longman, Harlow.Mayo, E.J. and Jarvis, L.P. (1981) The Psychology of Leisure Travel: Effective Marketing and Selling of Travel

Services. CBI Publishing Company, Boston.McLellan, R.W. and Fousher, K.D. (1983) Negative images of the United States as expressed by tour opera-

tors from other countries. Journal of Travel Research 22, 2–5.Morrison, A.M. (1989) Hospitality and Travel Marketing. Albany, New York.Peabody, D. (1985) National Characteristics. Cambridge University Press, Cambridge.Pearce, P.L. (1982) Perceived changes in holiday destinations. Annals of Tourism Research 9, 145–164.Pizam, A. (1999) Cross-cultural tourist behavior. In: Pizam, A. and Mansfeld, Y. (eds) Consumer Behavior in

Travel and Tourism. The Haworth Press, New York, pp. 393–411.Pizam, A. and Jeong G.H. (1996) Cross-cultural tourist behavior. Perceptions of Korean tour-guides.

Tourism Management 17, 277–286.Pizam, A. and Reichel, A. (1996) The effect of nationality on tourist behavior: Israeli tour guides’ percep-

tions. Journal of Hospitality and Leisure Marketing 4, 23–49.Pizam, A. and Telisman-Kosuta, N. (1989) Tourism as a factor of change: results and analysis. In:

Bystrzanowski, J. (ed.) Tourism as a Factor of Change: a Socio-Cultural Study, Vol. II. EuropeanCoordination Centre for Documentation in Social Sciences, Vienna.

Plog, S.C. (1990) A carpenter’s tools: an answer to Stephen L.J. Smith’s review of psychocentrism/allocen-trism. Journal of Travel Research 28, 55–58.

Reisinger, Y. (1997) Cross-cultural differences in tourism. Tourism Management 18, 139–147.Reisinger, Y. and Turner, L. (1997) Cross-cultural differences in tourism: Indonesian tourists in Australia.

Tourism Management 18, 139–147.Reisinger, Y. and Turner, L. (1998) Cross-cultural differences in tourism: a strategy for tourism marketers.

Journal of Travel and Tourism Marketing 7, 79–106.Richardson, S.L. and Crompton, J.L. (1988) Vacation patterns of French and English Canadians. Annals of

Tourism Research 15, 430–448.Ruiz, A.V., Olarte, R. and Iglesias, V. (1999) Evaluación de los destinos turísticos en función de su valor de

marca. In: Proceedings of the XI Encuentros de Profesores Universitarios de Marketing. Valladolid, Spain, pp.427–450.

Sirakaya, E., McLellan, R.W. and Uysal, M. (1996) Modelling vacation destination decisions: a behaviouralapproach. Journal of Travel and Tourism Marketing 5, 57–75.

Sussmann, S. and Rashcovsky, C. (1997) A cross-cultural analysis of English and French Canadians’ vaca-tion travel patterns. International Journal of Hospitality Management 16, 191–208.

You, X., O’Leary, J., Morrison, A. and Hong, G.S. (2000) A cross-cultural comparison of travel push andpull factors: United Kingdom vs. Japan. International Journal of Hospitality and Tourism Administration 1,1–26.

WTO (1997) International Tourism: a Global Perspective. World Tourism Organization, Madrid.WTO (2003) Tourism Highlights 2002. World Tourism Organization, Madrid.



Chapter twenty-sevenJourneys of the Imagination? The Cultural Tour RouteRevealed

Tove OliverInstitute of Rural Studies, University of Wales, Aberystwyth SY23 3AL, UK

Abstract

Organized tours are one of the main ways that tourists experience cultural destinations. They are oftendescribed as ‘a destination bubble’, conveying a sense of isolation rather than involvement. The extent towhich tour participants interact with and learn about destinations is not well understood, although theacquisition of knowledge is frequently cited as significant in peoples’ decisions to travel by this mode. Thischapter investigates cultural tour participants’ experiences, and specifically addresses the extent to whichparticipants’ images of their destinations change or remain unaltered after their visit, and whether satisfac-tion from a tour can be linked to the degree of informal learning gained about the route. The empiricalresearch employed route mapping to elicit information about tour members’ knowledge before and aftertouring. Ireland was selected because it presented elements common to many non-specialized cultural touritineraries. A multi-method approach combined qualitative and quantitative techniques in the analyses ofthe route maps, and triangulated the findings with those from focused interviews and participant observa-tion. The study found that tourists’ images changed in magnitude as the tour had enforced already well-defined images; it concludes that the potential for tour members to interact with host communities, lies notonly with the tourists themselves, but also with destination image intermediaries and tourism policy makers.

Introduction

Studies exploring the tour experience covera wide range of disciplinary influences,especially those from sociology and psychol-ogy, anthropology, geography and econom-ics. Approaches include: service-based andproduct-based satisfaction measurement(Geva and Goldman, 1991; Hughes, 1991),expectation and satisfaction measurement(Thomson and Pearce, 1980; Whipple and

Thatch, 1988; Geva and Goldman, 1989;Dunn-Ross and Iso-Ahola, 1991; Duke andPersia, 1994, 1996; Reisinger and Waryszak,1994; Yang, 1995) supplemented by observa-tion (Quiroga, 1990), and an evaluationusing pictorial images of the tourist setting(Chadee et al., 1996). While these evalua-tions focus on tour members’ motivations,expectations and experiences, they do notenable a comprehensive understanding oftourists’ engagement with their destina-



tions, nor an assessment of whether destina-tion images have changed or remain unal-tered. Other related bodies of literatureinclude ‘destination itineraries’ (Forer andPearce, 1984; Weightman, 1987; Mings andMcHugh, 1992; Dahl, 1999; Oppermann,2000), although, like the research intotourism’s promotional literature (Goodhalland Bergsma, 1990; Holloway andRobinson, 1995; Dann, 1996, 2000), thesestudies tend to provide insights into opera-tors’ perceptions of destinations, ratherthan providing customer-focused evalua-tions (Oppermann, 1995).

According to Crompton (1979), the imageof a tourist destination is the collective sum ofa tourist’s beliefs, ideas, impressions andexpectations. The direct experience of a des-tination is likely to provide the bulk of thetour experience (Hanefors and Larsson-Mossberg, 1998). Many studies have used theanalogy of a tourist or environmental ‘bub-ble’ to describe this temporal and spatialspan, enforcing the idea that this is a ‘blackbox’ in terms of academic understanding ofthe consumption processes involved (Oliver,2001a,b). An investigation of tourists’ under-standing of transient destination imagesmakes the concern one of environmentalcognition (Evans, 1980; Spencer et al., 1989).This field is extremely fertile in methods toelicit both recall and/or recognition of envi-ronments (Moore and Golledge, 1976) andapproaches have assessed the responses oftravellers as they move across a landscape(Appleyard et al., 1964; Carr and Schissler,1969). The technique of cognitive mapping,while having an extensive theoretical andtechnical grounding in environmental psy-chology (Lynch, 1960) and human geography(Gould and White, 1974), has been relativelyunexplored in the field of tourism, althougha small body of literature has investigatedtourists’ spatial learning in unfamiliar envi-ronments (e.g. Pearce, 1977; Walmsley andJenkins, 1992; Page, 1997). A cognitive sketchmap, pioneered by Lynch (1960) in the semi-nal work The Image of the City, is the instru-ment by which spatial knowledge can beexternalized. A linear form of the cognitivemap is the ‘strip map’ or ‘route map’ (Pearce,1981; Bell, 1999). These have been used for

many centuries, predominantly to representroutes between a place of origin and a desti-nation, and a variation was used by theancient Egyptians to depict the route thedead would follow to ‘the beyond’ (Yaru).However, these techniques have never beenused to explore the perceptions of touriststraversing large-scale environments in com-mercially organized groups (Pearce, 1981;Bell, 1999).

Methodology

A cultural tour can be described as a mobileexperience of a cultural destination, usuallyby land or water, along trajectories that linkcultural attractions. The objective of the studywas to explore the extent to which partici-pants’ images of these tour routes changed orremained unaltered after they had experi-enced them. It further aimed to assesswhether participants’ levels of informal envi-ronmental learning gained from a tour, cor-related with the levels of satisfaction that theyexperienced. Tourists’ motivations to under-take tours, and their resulting levels of satis-faction, vary according to tour segment (i.e.whether non-specialized or niche-based).This study was therefore confined to non-spe-cialized tours for international markets(Stebbins, 1996).

Judgement and convenience-based sam-pling were used to select a cultural destina-tion and Ireland was chosen because itpresented elements common to many non-specialized tour itineraries in Europe. A lon-gitudinal, survey-based research strategymeasured respondents’ knowledge and atti-tudes before and after touring. The samplingframe included all first-time visitors partici-pating in a 12-day or less, non-specializedtour of Ireland. The survey was administeredin two parts: the pre-tour questionnaireswere interviewer distributed at DublinAirport (23–30 September 1999), and follow-up post-tour questionnaires were sent by air-mail to participants’ homes. In addition, 50post-tour questionnaires were distributed inthe departures areas of Dublin Airport as itwas anticipated that less post-tour formswould be returned.

320 T. Oliver


The questionnaires consisted of two mainparts, separating the predominantly qualitativemap element from the mainly quantitative atti-tude measurement and subject traits. A rangeof response formats was used, including open-ended and closed questions; an outline basemap of Ireland; and prompts in the form ofpredetermined symbols developed from themapping language ‘Environmental A’ (Woodand Beck, 1975). These devices were intendedto encourage both depth and breadth of spatialexpression. If the mapping language had notbeen given, it is likely that participants wouldhave simply referred to ‘factual’ data, followingtraditional perceptions of cartography.

The explanatory potential of cognitivemapping can be limited, and previousresearch has indicated that it could benefitfrom a multiple method perspective (Guy etal., 1990). Therefore, qualitative informationgained through interviews with elite infor-mants, eliciting specialist knowledge aboutthe touring destination and by participantobservation on a scheduled tour (Oliver,2000), was reassessed through methodologi-cal and data triangulation with the surveymaterial (Oliver, 2002; also see Decrop,1999). The elite informants sample includedexecutives in Bord Fáilte (the NationalTourism Organisation), and in HeritageIsland who promote Ireland’s heritage; acade-mics specializing in Ireland’s cultural land-scapes; and the manager of a British-basedtour company operating in Ireland. In adopt-ing an integrative strategy, the aim was to bal-ance the strengths and weaknesses of thesethree approaches and to overcome inherentbiases (McIntosh, 1998).

Response rates

Responses were elicited from most of the 19products surveyed, over two-thirds of whichyielded pre-tour and post-tour responses fromthe same individuals. In total, 110 valid ques-tionnaires were returned; 63 (57%) were pre-tour questionnaires and the remainder (47;53%) were post-tour questionnaires. There wasa 65% response rate to the postal survey and a40% and a 20% response rate to each of thesurveys that were administered by the inter-

viewer. The low response rate (20%) from the‘captive’ post-tour audience approached in thedepartures areas contradicts the apparententhusiasm with which these questionnaireswere accepted; many potential participantsclaimed that it would give them something todo on their in-bound transatlantic flights (Day,1999). The timing of this survey could havebeen a contributory factor (O’Neill, 1998;Turley, 1999), and this could explain the rela-tively high response (65%) from the postal sur-vey when tour members had returned home;several of these respondents commented thattogether with their photographs, the timing ofthis survey had helped them to ‘relive’ theirtouring experiences.

Analyses and Findings

A cornucopia of images, bewildering in theirvariety: this is the world of maps

The Power of Maps (Wood, 1992, p. 4).

The aim of the analyses was to understand thecontent of the maps rather than to assess theveracity and orientation of participants’ spatialknowledge (Walmsley and Jenkins, 1992).Lynch (1960) proposed that individuals learnabout complex environments by building upsimplified images. In his view, the processfocuses on five elements of the environment:paths are the routes along which people move;edges are obstacles or lines separating differ-ent parts of the landscape; nodes are placesthat serve as foci for travel; districts are rela-tively large areas with an identifiable character;and landmarks are points of reference used innavigation. Paths, landmarks and districts arethe three features used most often in city map-ping research (Walmsley and Jenkins, 1992).Pearce (1981) also used these categories in hisstudy of route maps in a rural context,together with concepts he adopted from out-side the cognitive mapping arena; ‘texture’referring to a general non-spatial commentary(Spreiregen, 1965, Fairbrother, 1972, cited inPearce, 1981, p. 145) and a social scoredescribing social activities (Lee, 1968, cited inPearce, 1981, p. 145). The score for any cate-gory is simply the frequency of that item(Pearce, 1981; Walmsley and Jenkins, 1992).

The Cultural Tour Route Revealed 321


Map components

In this study, the maps in both the pre-tourand post-tour questionnaires were visuallycompared. Three of the basic componentsselected by participants to express spatialknowledge of their touring destination resem-bled those used in previous mapping research(also see Beck and Wood, 1976; Wall, 1997):

• Points (or landmarks) are a uni-dimen-sional site or place with a suggested, evenif inaccurate, location.

• Lines (or paths) are the representation ofa tour route or road system, and are essen-tially two dimensional.

• Areas (or districts) are regions exhibitingparticular characteristics, although they donot need to have a clearly defined boundary.

• A fourth category included exampleswhere knowledge about the destinationcould not be described in spatial terms. Inthese examples, the map was left com-pletely blank; symbols were drawn outside

the map outline; or were drawn within themap outline but apparently in a com-pletely random fashion (Fig. 27.1 a–c).

Sample profile

About two-thirds (67%) of the survey samplewas female, and a third male. Most (83%)were US residents, with others residing inBritain (8%), Australia (3%) and Canada(2%). The composition of travelling compan-ions was consistent with those of the partici-pant observation study (Oliver, 2000), withalmost three-quarters (71%) travelling withtheir partner or spouse and only 5% travel-ling alone. Only 3% were experienced tourmembers, having already toured more thanseven times, while almost half (48%) had nottoured by coach before, contradicting theperception that there is a common ‘type’ ofgroup traveller. In addition, over half of thesample (51%) preferred to explore alone or

322 T. Oliver

Fig. 27.1. (a) Route maps dominated by points; (b) route maps dominated by lines; (c) route mapsdominated by areas (left) and non-spatial representations (right).



Fig. 27.1. Continued.


to socialize in a small group whilst on holiday;only 19% favoured a large group. Over athird (40%) of the respondents were between55 and 64 years old and many were to eitherside of this age bracket.

Motivations

The main reason for choosing a tour was tosee the highlights of Ireland (42% of partici-pants selected this as a motivation). This wasfollowed by: good value for money (32%);that information was provided (31%); safetyand security (26%); comfort (26%); sociabil-ity (23%); to get ideas for an independenttrip (19%); time constraints (11%); specialhealth needs (5%); with 15% citing anothermain reason for selecting a tour, including13% who said that not having to drive was animportant factor (Fig. 27.2).

Aggregate content analyses

Aggregate cognitions represent a consensualview of a place and provide a statistically supe-rior basis on which to draw conclusions (Beckand Wood, 1976). The aggregate sample con-sisted of 110 valid responses. Content analysisenabled the textual and pictorial materialfrom the route maps to be comparedtogether and this seems to have overcomesome of the problems associated with inter-preting ‘drawings’ and their intended mean-

ings (for example, Gamradt, 1995). Contentanalysis allows the systematic analysis of non-statistical material (Finn et al., 2000). It per-mits inferences to be made about the senderof the message, the message or the audience(Krippendorff, 1980; Werber, 1990),although it is frequently used implicitly in theanalysis of cognitive maps.

The content analyses of the route mapsconsisted of two main activities: the transcrip-tion of all textual material relating to map con-tent, and the scoring of all pictorial symbolsand their descriptive labels. Scores werederived from themes that participants felt weresignificant in their touring experiences. Thesymbols and their text labels were quantifiedby entering each score into a database createdin SPSS for Windows (Release 10.0). The scorefor any category is simply the frequency of thatitem. A reliability check was built into themethodology by re-analysing the material 1month later, producing a level of agreement ofapproximately 90% (Finn et al., 2000).

The responses were weighted to takeaccount of the different response ratesbetween the pre-tour and post-tour surveys,before comparing the aggregate scores fromtheir map content. Descriptive statistics com-puted using SPSS revealed that map contentincreased after touring; both the number ofparticipants selecting symbols and the fre-quency with which symbols were used in a sin-gle map were greater than in the pre-tourversions. However, it was important to deter-mine whether these observed changes werestatistically significant.

Testing for significant relationships

Although the survey sample was suitablyrobust, the data collected through the contentanalyses could not be assumed to be normallydistributed, because some of the content cate-gories had very low scores. It was thereforedecided to use the non-parametric alternativeto the independent-samples t-test, theMann–Whitney U test, to test whether one sur-vey population had larger values than theother (Norušis, 2000). This procedure is usedwhen testing between two independentgroups when the assumptions for the paramet-

324 T. Oliver

Participants' motivations

ValueTime

SociabilitySafety

OtherInformation

IdeasHighlights

HealthComfort

Always

Num

ber

of p

artic

ipan

ts

50

40

30

20

10

0

Fig. 27.2. Participants’ motivations.


ric t-test cannot be met. It is distribution free,but it requires that certain assumptions bemet (Wonnacott and Wonnacott, 1990; Burns,2000). It must be possible to rank the scoresproduced by individuals, and it is thereforeideally suited to the data collected by this sur-vey. The mathematics of this test is based onthe simple observation that if there is a realdifference between scores in two samples thenthe scores in one should be generally largerthan the scores in the other sample.

The Mann–Whitney U test was computedusing SPSS for the content scores of the pre-tour and post-tour surveys for the aggregatesample, where:

H0: the null hypothesis states that the twopopulations are identical with respect totheir map content values.

H1: the alternative hypothesis states thatthe two populations are not identical withrespect to their map content values.

The null hypothesis is that the two popula-tions are identical, so rejection could bebecause their means, variances or the shapeof the distributions differ significantly (Silver,1992). The Mann–Whitney U test, whenapplied to the scores of tour images, foundthere was a significant difference between thepre-tour and post-tour scores for the aggre-gate sample (U = 4390.000, P = 0.000), withpost-tour scores significantly greater. The nullhypothesis is therefore rejected at the 95%confidence interval.

Assessing the strength of the relationship

A Spearman rank correlation coefficientenables the strength of the relationshipbetween two ranked groups to be quantified.The type of data to which a rank correlationcoefficient is particularly suited is that in whichneither of the variables represent a characteris-tic that can be precisely measured by an objec-tive standard (Letchford, 1986). This isespecially true when a margin of subjective dif-ference between scores is possible, even thoughthere has been an agreed standard of marking.These subjective differences will tend to be lesspronounced among the ranks of marks thanbetween the scores themselves, and so a rank

correlation measurement is less susceptible tosubjective bias than, for example, a product-moment coefficient (Letchford, 1986).

The correlation between the map contentfrom the pre-tour and post-tour maps wascomputed by SPSS using the Spearman’s rankcorrelation coefficient. This shows there is astrong positive relationship between the num-bers from each content category. The correla-tion between the pre-tour and post-touraggregate sample is strongly positive at 0.727.The P-value 0.000 indicates the probability ofthis correlation occurring by chance is lessthan 0.1 (10%; two-tailed). This correlation istherefore significant at the 95% level of confi-dence. However, given the extremity of thesingle value representing ‘places’, it wasimportant to ascertain whether this point,although valid, was distorting the results(Norušis, 2000). The Spearman’s coefficientwas therefore recalculated after first remov-ing this value from the dataset. The r-value(0.720) demonstrates a strong positive corre-lation between the pre-tour and post-tourdata, significant at the 0.1 (10%) level ofprobability, without the outlying value.

Altered images?

Taken together, the Mann–Whitney U and theSpearman’s rank correlation coefficient pro-vide highly complementary findings, suggest-ing that although there is a significantdifference between the magnitude of destina-tion images represented before and after tour-ing, the samples are related in terms of theorder and importance of images depicted. Thissuggests that the pre-tour images were strength-ened by the tour experience, rather thanchanged by it. This is particularly well illus-trated by Fig. 27.3 where there is very littlechange in the categories comprising more than2% of the total image, before and after touring.

Overall, the change in magnitude ofimages depicted before and after touring wassignificant (above). Several items in particu-lar scored very highly, most notably placeswhere 68% of pre-tour participants and 87%of post-tour participants made references tocities, towns or villages, usually by drawingspecific points on the maps.



The tour route itself was the most signifi-cant linear element depicted on the post-tour maps; explicit representations weremade by labelling or indicating the direc-tions of travel (37% of the pre-tour sampleand 53% of the post-tour sample referencedthis category). There are instances where thesame participant changed the style of linefrom, for example, a dashed line to a hardline after touring. This appears to demon-strate the greater confidence with which indi-

viduals could represent their journey, but itdoes not conclusively support a staged modelof spatial learning, such as proposed byanchor point theory, whereby locations aresubsequently linked by routes and finally sur-rounded by areas (Walmsley and Jenkins,1992). Most of the tour participants had animpression of their itinerary before touring,and both their linear images and placeimages were either maintained or strength-ened by the visit.

326 T. Oliver

Other

Sightseeing

Sacred sites

Roads

Ring of Kerry

Positive

Places

Photography

Pubs

Friendly

Cliffs

Castles

Beautiful in west

Beautiful

Other

Sacred sites

Ring of Kerry

Positive

Places

Photography

Memorable

Hotels

Pubs

Friendly

Castles

Beautiful

Fig. 27.3. Images from the aggregate sample: pre-tour (top) and post-tour (below).


Satisfaction ratings

Several studies have explored the factorsrelating to customers’ levels of satisfactionwith their tours, although these have notincluded tourists’ knowledge about theirdestinations. Participants were asked to ratetheir overall satisfaction levels, and to assesswhether their experiences matched theirexpectations. The majority (81%) said thatthey had a very good experience and 60%said that the tour exceeded their expecta-tions. The mean scores of the mapped sym-bols from the post-tour survey were thencalculated according to the satisfaction rat-ings (n = 47). This found quite similar scoresbetween the different groups, although,interestingly, the groups who said that theyhad a very good experience and those forwhom the experience had exceeded theirexpectations provided more detail in theirmaps. The mean scores for the satisfactionratings were as follows: okay (x

_= 21); good

(x_

= 21); very good (x_

= 23); below expecta-tions (x

_= 21); matched expectations (x

_=

21); exceeded expectations (x_

= 24).The Mann–Whitney U test was applied to

the content scores of two groups, those whohad a good experience and those who had avery good experience where:

H0: the null hypothesis states that the twopopulations are identical with respect totheir map content values.

H1: the alternative hypothesis states thatthe two populations are not identical withrespect to their map content values.

This found that there was no significant dif-ference between them (U = 139.000, P =0.706). The test was repeated for the groupsfor whom their expectations were matchedand for those for whom their expectationswere exceeded. This also found that therewas no significant difference between them(U = 189.500, P = 0.601) at the 95% level ofconfidence. While these findings indicate alink between high levels of satisfaction and agreater knowledge of the touring environ-ment, they are not significant statistically atthe 95% level of confidence. One of the keymotivations that these tourists had for select-

ing a cultural tour was to gain informationabout the destination; this suggests that theywould actively seek information and wouldbe more satisfied if they were successful.

The tour members were asked whetherthere was anything that their tour did notsufficiently provide (n = 47) – 28% of partici-pants said that they felt that their tourslacked in ideas for an independent trip; 23%said that they would have liked more time atplaces; 17% said that they would have pre-ferred more variation in the itinerary; and15% said that they felt that their experiencelacked authenticity; 13% said that their tourdid not provide adequately for their specialneeds (such as health and dietary require-ments); and only 4% said that they thoughtthat their tour did not provide sufficientinformation, thus implying that the otherparticipants were satisfied with the amount ofinformation they received (Dunn-Ross andIso-Ahola, 1991; Fig. 27.4).

The tourists’ qualitative descriptions pro-vide further insight into these dissatisfactionscores. For example, some participants wouldhave liked the time and opportunity to pene-trate further into ‘authentic’ culture and toidentify with local experience:

Too much time in busToo much herding to pre-arranged rest stopsNo evening entertainmentNot enough stopping time in non-tourist shops

(Female, New Jersey, USA)

The pace was fast – it would have been nicer toget to towns at 3.00 and spend time with thepeople and culture, i.e. more independenttime, less commercial

(Male, New York, USA)

Several participants felt more able to travelindependently in Ireland as a result of theirguided tour:

Now I know a little about it, I’d be happy todrive and do B/B’s!

(Female, New South Wales, Australia)

Now I know where to go, I can do it at my ownpace. The pace of tour was quite fast

(Female, Michigan, USA)

The fast pace of tours was a major causefor dissatisfaction, although tourists wereaware that a trade-off had occurred between



this and the number of highlights included inan itinerary:

It would have been nice to have more time atplaces but then we would not have seen asmuch

(Female, Minnesota, USA)

Conclusions

Through images and texts, attempts are madeto attract tourists to rural areas through the pro-motion of representations of idealized, sym-bolic, cultural landscapes

Rural Cultural Economy: Tourism and SocialRelations (Kneafsey, 2001, p. 762).

The study set out to identify whether and pre-cisely how tourists’ perceptions of organizedcultural tours change as a result of theirdirect experience of their touring environ-ments, namely their tour routes. This objec-tive has been successfully accomplished byfinding that tourists’ images changed in mag-nitude: the tour experience enforced prede-fined images, rather than generating newones. Overall, tour members’ pre-tour routemaps were quite well advanced. The most rea-sonable explanation for this seems to be thatsignificant learning about the tour routeoccurred in the pre-purchase and planningstages of the trip. Image is a key marketingtool in an industry where potential con-

sumers must base buying decisions uponmental images of product offerings ratherthan being able physically to sample alterna-tives (Morgan and Pritchard, 1998). Peopleare therefore likely to compare different tourproducts and assess their ‘value for money’before making a final selection. This wouldexplain why knowledge of places and routesexisted at an early stage of the tour experi-ence. Future work focusing on the criticalpre-purchase period in the destination-imageformation process could make a significantcontribution to this area of research.

The most significant learning was in theplace category, and places tended to be posi-tioned along a discernible route. The tran-sient nature of the tour experience is likely todiffer considerably from one in which envi-ronmental knowledge is built up over aperiod of days or even weeks. However, thefinding that knowledge of places positionedconsecutively along a route tends to increase,indicates a sequential rather than a spatialform of learning. This corroborates work bySpencer and Weetman (1981) who found intheir study of new residents in Sheffield, UK,that where a task is to represent a simpleroute, the tendency is to select a sequentialstyle of mapping. Only when the task is con-siderably more complex do individuals movefrom an initial sequential mode to a devel-oped spatial image.

328 T. Oliver

Other

Visits

Variation

Time

Special interest

Practical solutions

SociabilitySpecial needs

Information

Ideas

Value

Organization

Comfort

Authenticity

Fig. 27.4. Attributes that were not sufficiently provided on tour.


Cultural tour participants’ pre-tour mapsof Ireland reflected imagery used by BordFáilte and tour operators in the marketing ofthis destination, as revealed by theexploratory interviews (Oliver, 2002). Thekey themes depicted are ‘scenery’, ‘people’and ‘the past’. It seems that generic, stereo-typical images of a traditional rural life and ahomogenous ‘Heritage Island’ were retainedbecause these images determined the successof the tour. Williams (1998) has drawn atten-tion to this phenomenon; he says that boththe providers of tourism construct specificimages to attract the visitor, while the touristmust in turn confirm his or her expectations:

In this way, tourist images tend to become self-perpetuating and self-reinforcing with the atten-dant risk that, through time, tourist experiencesbecome increasingly artificial.

(Williams, 1998, p. 178)

Greenfield (2000) notes, ‘our visual expe-rience, is a kind of mixture of informationcoming from our eyes and prior associations– how else might we interpret what we seeand give the world significance?’ (p. 74).This notion of a ‘mind’s eye’ helps to explainwhy tour members continued to focus onimages which reflect an ‘Irish idyll’ (forexample, beautiful scenery, castles, cliffs,mountains, lakes, sheep, friendly people andgoing to the pub) and specific products (forexample, Guinness, Waterford Crystal andBelleek China). Tourists’ images tend to haveconsiderable stability over time (Gunn, 1972;Fakeye and Crompton, 1991), and they canplay an active part in shaping the perceivedperformance of a tour; if destination imagesare realized then tourists are more likely toenjoy the experience (Geva and Goldman,1991). This helps to explain why images aremaintained long after the factors that havemoulded them have changed. Even in theface of dramatic changes in destinationattributes, such as a recent transition inIreland’s economy, there is no significantchange in the perception of place, as touristand facilitator have actively endorsed theanticipated images.

The degree to which tourists engage withand are engaged by their host environment,has interested tourism researchers fordecades; Urry’s ‘tourist gaze’ (Urry, 1990) is a

process by which tourists seek authenticity andtruth in places away from their own everydaylife. MacCannell (1976) sees tourists as aspir-ing to what he terms ‘back-room’ experiences,while other writers consider that ‘soft’ tourismpractices have a greater opportunity to pene-trate further into local culture (Smith, 1989).‘Soft’ tourism is typically more individuallybased than the tours segment, and centred ona sense of place involving local products andcommunities, with respect for local environ-ments. Tourists engaged in soft tourism activi-ties are likely to be less accepting of‘pseudo-events’ and manufactured experi-ences, and more concerned with ‘authentic’experiences (Jenkins and Oliver, 2001).Ironically, high levels of satisfaction can beachieved when tourists feel that they haveceased to be tourists, and assumed the role of‘guests’ (Ryan, 1991). The apparent conflictbetween the medium of the cultural tour,which tends to isolate tourists from culturalenvironments, and the actual product, localculture, has been highlighted by this study.This creates an image of ‘hard’ tourism, whichtends to exhibit little penetration into hostcommunities, with links tending to be interna-tional rather than local (Stabler, 1997).

This study determined two key factorswhich can potentially lead to tour members’dissatisfaction: the fast pace of the tours, andthe lack of ideas provided by tour personnelor the tour company to enable a return visit.The pace of a tour will be a function of thenumber of highlights included in the itineraryand it will help to establish whether a tour is‘good value for money’; both are key motiva-tions for taking a tour. In recent years, touroperators have acknowledged a market fortours that include ‘extra leisure’ time, and itwill be interesting to see whether this productvariation is successful, especially with first-timevisitors. Other approaches include the use of‘radial’ touring patterns where a tour is basedin one location and then different attractionsare explored from this centre (SeamusCaulfield, Dublin, 1999, personal communica-tion). This approach provides an alternativeto conventional tour circuits (Forer andPearce, 1984) and offers both extra leisuretime for tourists and opportunities for themto meet local people. A more ‘sedentary’ tour



could bring economic benefit to local regions,particularly in less well-known touring areas,and route mapping offers a useful means toexplore such itineraries.

Several participants in this study indicatedthat the guided tour experience had giventhem the confidence to return to Irelandindependently; it appears that these partici-pants could potentially ‘progress’ in theirtourism behaviours, consistent with the ideaof Pearce’s travel career ladder (Pearce,1988). Pearce (1988) postulated the existenceof a ‘travel career ladder’ where, as touristsbecome more experienced, so they becomemore adventurous and curious about otherplaces, and eventually they travel indepen-dently. Despite this, almost a third of therespondents said that they had not been pro-vided with sufficient information to enablethem to do so. This finding indicates that des-tination managers and tour operators (offer-ing ‘self-drive’ itineraries) could encourage

independent visits by targeting first-timetourists travelling in organized groups; how-ever, the precise stage at which consumers aremost receptive needs to be determined.

The potential for tour members to movebeyond experiencing images, and eventuallyparticipate with and contribute to local com-munities, lies not only with the tourists them-selves, but also with destination imageintermediaries and tourism policy makers.This study began by questioning whether thecultural tour isolates tourists from their hostenvironments. It has successfully answeredthis and concludes by asking, ‘what are thealternatives?’

Acknowledgements

The author is grateful to the Revd Prof. MyraShackley, Dr Christopher Spencer and Dr JaneBinner, for their comments on an earlier draft.

330 T. Oliver

References

Appleyard, D., Lynch, K. and Meyer, J.R. (1964) The View from the Road. MIT Press, Massachusetts Instituteof Technology, Cambridge, Massachusetts.

Beck, R. and Wood, D. (1976) Comparative developmental analysis of individual and aggregated cognitivemaps of London. In: Moore, G. and Golledge, R. (eds) Environmental Knowing. John Wiley & Sons,New York.

Bell, S. (1999) Cartographic presentation as an aid to spatial knowledge acquisition in unknown environ-ments. PhD thesis, UCSB, Santa Barbara, USA.

Burns, R. (2000) Research Methods. Sage, London.Carr, S. and Schissler, D. (1969) Perceptual selection and memory in the view from the road. Environment

and Behaviour 1, 7–35.Chadee, D., Doren, D. and Mattsson, J. (1996) An empirical assessment of customer satisfaction in tourism.

The Service Industries Journal 16, 305–320.Crompton, J.L. (1979) An assessment of the image of Mexico as a vacation destination and the influence

of geographical location upon image. Journal of Travel Research 17, 18–23.Dahl, R. (1999) Tracing complex travel behaviour. In: Ruddy, J. (ed.) Tourism Destination Marketing. Gaining

the Competitive Edge. European Conference of Travel and Tourism Research Association, 29September–2 October 1999. TTRA, Dublin, pp. 1–16.

Dann, G.M.S. (1996) The people of tourist brochures. In: Selwyn, T. (ed.) The Tourist Image. Myths and MythMaking in Tourism. John Wiley & Sons, Chichester, pp. 61–81.

Dann, G.M.S. (2000) National tourist offices and the language of differentiation. In: Gartner, W.C. andLime, D.W. (eds) Trends in Outdoor Recreation, Leisure and Tourism. CAB International, Wallingford,UK, pp. 335–345.

Day, G. (1999) Trainspotting: data collection from a captive audience. Tourism Management 20, 153–155.Decrop, A. (1999) Triangulation in qualitative tourism research. Tourism Management 20, 157–161.Duke, C.R. and Persia, M.A. (1994) Foreign and domestic escorted tour expectations of American

travelers. Journal of International Consumer Marketing 6, 61–77.


Duke, C.R. and Persia, M.A. (1996) Consumer-defined dimensions for the escorted tour industry segment:expectations, satisfactions, and importance. Journal of Travel and Tourism Marketing 5, 77–99.

Dunn-Ross, E.L. and Iso-Ahola, S. (1991) Sightseeing tourists’ motivation and satisfaction. Annals ofTourism Research 18, 226–237.

Evans, G.W. (1980) Environmental cognition. Psychological Bulletin 88, 259–287.Fakeye, P.C. and Crompton, J.L. (1991) Image differences between prospective, first-time and repeat visi-

tors to the Lower Rio Grande Valley. Journal of Travel Research 30, 10–16.Finn, M., Elliot-White, M. and Walton, M. (2000) Tourism and Leisure Research Methods: Data Collection,

Analysis, and Interpretation. Longman, Harlow, UK.Forer, P.C. and Pearce, D.G. (1984) Spatial patterns of package tourism in New Zealand. New Zealand

Geographer 40, 34–42.Gamradt, J. (1995) Jamaican children’s representations of tourism. Annals of Tourism Research 22, 735–762.Geva, A. and Goldman, A. (1989) Changes in the perception of a service during its consumption: the case

of organized tours. European Journal of Marketing 23, 44–52.Geva, A. and Goldman, A. (1991) Satisfaction measurement in guided tours. Annals of Tourism Research 18,

177–185.Goodhall, B. and Bergsma, J. (1990) Destinations as marketed in tour operators’ brochures. In: Ashworth,

G. and Goodall, B. (eds) Marketing Tourism Places. Routledge, London, pp. 170–191.Gould, P. and White, R. (1974) Mental Maps. Penguin Books Ltd, Harmondsworth, UK.Greenfield, S. (2000) Brain Story. BBC Worldwide Limited, London.Gunn, C.A. (1972) Vacationscape: Designing Tourist Regions. Bureau of Business Research, University of

Texas, Austin, Texas.Guy, B.S., Curtis, W.W. and Crotts, J.C. (1990) Environmental learning of first-time travellers. Annals of

Tourism Research 17, 419–431.Hanefors, M. and Larsson-Mossberg, L. (1998) The tourism and travel consumer. In: Gabbott, M. and

Hogg, G. (eds) Consumers and Services. John Wiley & Sons, Chichester, UK, pp. 141–161.Holloway, J. and Robinson, C. (1995) Marketing for Tourism. Longman Group Ltd, Harlow, UK.Hughes, K. (1991) Tourist satisfaction: a guided ‘cultural’ tour in North Queensland. Australian Psychologist

26, 166–171.Jenkins, T.N. and Oliver, T. (2001) Integrated Tourism: A Conceptual Framework. Institute of Rural Studies,

University of Wales, Aberystwyth.Kneafsey, M. (2001) Rural cultural economy: tourism and social relations. Annals of Tourism Research 28,

762–783.Krippendorff, K. (1980) Content Analysis: an Introduction to its Methodology. Sage, Newbury Park, California.Letchford, S. (1986) Business Mathematics and Statistics. Cassell, Eastbourne, UK.Lynch, K. (1960) Image of the City. MIT Press, Cambridge.MacCannell, D. (1976) The Tourist: a New Theory of the Leisure Class. Macmillan, London.McIntosh, A. (1998) Mixing methods: putting the tourist at the forefront of tourism research. Tourism

Analysis 3, 121–127.Mings, R.C. and McHugh, K.E. (1992) The spatial configuration of travel to Yellowstone National Park.

Journal of Travel Research 30, 38–46.Moore, G.T. and Golledge, R.G. (eds) (1976) Environmental Knowing. Theories, Research, and Methods. John

Wiley & Sons, New York. Morgan, N. and Pritchard, A. (1998) Tourism Promotion and Power. Creating images, creating identities. John

Wiley & Sons, Chichester, UK.Norušis, M.J. (2000) SPSS 10.0 Guide to Data Analysis. Prentice-Hall Inc., Englewood Cliffs, New Jersey.Oliver, T. (2000) Watch this space: observing patterns of tourist behaviour on a cultural tour. In: Robinson,

M., Long, P., Evans, N., Sharpley, R. and Swarbrooke, J. (eds) Motivations, Behaviour and Tourist Types.Tourism 2000: time for a celebration?, Sheffield, 2–7 September 2000. Centre for Travel and Tourismand Business Education Publishers Ltd, Sunderland, pp. 321–330.

Oliver, T. (2001a) Consuming culture: the case of the cultural tour route. Tourism Analysis 5, 177–182.Oliver, T. (2001b) The consumption of tour routes in cultural landscapes. In: Mazanec, J.A., Crouch, G.I.,

Brent Ritchie, J.R. and Woodside, A. (eds) Consumer Psychology of Tourism, Hospitality and Leisure, Vol. 2.CAB International, Wallingford, UK, pp. 273–284.

Oliver, T. (2002) Altered images? The case of the cultural tour route. PhD thesis, Nottingham TrentUniversity, Nottingham, UK.

O’Neill, M.A. (1998) The effects of survey timings on perceptions of service quality. Managing ServiceQuality 8, 126–132.



Oppermann, M. (1995) Comparative analysis of escorted tour packages in New Zealand and NorthAmerica. Progress in Tourism and Hospitality Research 1, 85–98.

Oppermann, M. (2000) Where psychology and geography interface in tourism research and theory. In:Woodside, A.G., Crouch, G.I., Mazanac, J.A., Oppermann, M. and Sakai, M.Y. (eds) ConsumerPsychology of Tourism, Hospitality and Leisure. CAB International, Wallingford, UK, pp. 19–37.

Page, S. (1997) Urban tourism: analysing and evaluating the tourist experience. In: Ryan, C. (ed.) TheTourist Experience. Cassell, London/New York, pp. 112–135.

Pearce, P.L. (1977) Mental souvenirs: a study of tourists and their city maps. Australian Journal of Psychology29, 203–210.

Pearce, P.L. (1981) Route maps: a study of travellers’ perceptions of a section of countryside. Journal ofEnvironmental Psychology 1, 141–155.

Pearce, P.L. (1988) The Ulysses factor. Evaluating Visitors in Tourist Settings. Springer-Verlag, New York.Quiroga, I. (1990) Characteristics of package tours in Europe. Annals of Tourism Research 17, 185–207.Reisinger, Y. and Waryszak, R. (1994) Japanese tourists’ perceptions of their tour guides: Australian experi-

ence. Journal of Vacation Marketing 1, 28–40.Ryan, C. (1991) Recreational Tourism, a Social Science Perspective. Routledge, London.Silver, S. (1992) Business Statistics. McGraw-Hill Book Company, London.Smith, V.L. (ed.) (1989) Hosts and Guests: the Anthropology of Tourism. University of Pennsylvania Press,

Philadelphia, Pennsylvania.Spencer, C., Blades, M. and Morsley, K. (1989) The Child in the Physical Environment. The Development of

Spatial Knowledge and Cognition. John Wiley & Sons, Chichester, UK.Spencer, C. and Weetman, M. (1981) The microgenesis of cognitive maps. Transactions of the Institute of

British Geographers 6, 375–384.Stabler, M.J. (1997) Tourism and Sustainability. CAB International, Wallingford, UK.Stebbins, R.A. (1996) Cultural tourism as serious leisure. Annals of Tourism Research 23, 948–950.Thomson, C.M. and Pearce, D.G. (1980) Market segmentation of New Zealand package tours. Journal of

Travel Research 19, 3–6.Turley, S. (1999) A case of response rate success. Journal of Market Research Society 41, 301–309.Urry, J. (1990) The Tourist Gaze. Leisure and Travel in Contemporary Society. Sage, London.Wall, G. (1997) Tourism attractions: points, lines, and areas. Annals of Tourism Research 24, 240–243.Walmsley, D.J. and Jenkins, J.M. (1992) Tourism cognitive mapping of unfamiliar environments. Annals of

Tourism Research 19, 268–286.Weightman, B.A. (1987) Third world tour landscapes. Annals of Tourism Research 14, 227–239.Werber, R.P. (1990) Basic Content Analysis. Sage, Newbury Park, California.Whipple, T.W. and Thatch, S.V. (1988) Group tour management: does good service produce satisfied cus-

tomers? Journal of Travel Research 27, 16–21.Williams, S. (1998) Tourism Geography. Routledge, London/New York.Wonnacott, T.H. and Wonnacott, R.J. (1990) Introductory Statistics for Business and Economics. John Wiley &

Sons, New York. Wood, D. (1992) The Power of Maps. Routledge, London.Wood, D. and Beck, R. (1975) Talking with Environmental A, an experimental mapping language. In:

Moore, G. and Golledge, R. (eds) Environmental Knowing. John Wiley & Sons, New York, pp. 351–361.Yang, M. (1995) An exploratory analysis of the travel benefits sought, travel satisfaction, culture shock, and

image differences among international tourists on bus tours between Los Angeles and Las Vegas(California, Nevada). PhD thesis, The Pennsylvania State University, Pennsylvania.

332 T. Oliver


Index

333

activities 57, 80–85, 92, 96activity-based segmentation 247activity-based models 91–92, 94, 100, 112choice of 164patterns of 109

advertising 69affection 277airlines 21–30, 237–243analytical methods 89,92ANOVA 220, 268, 272, 309–311, 315anticipation 22, 24arousal 22, 282attitudes 191–192Australia 76–90, 135–145, 169–178, 218Austria 11–18, 37, 39, 247–252

Bayesian statistics 98behavioural consistency 277budget motivations 58bundling 183–184

career choice 216CHAID 105–106, 108–109chi-squared analysis 138–139choice 92, 96–97, 98, 103, 121, 130, 169–178

of convention site 135–145of destination 103, 132, 135–145, 162, 170,

181–183, 187, 234, 265, 278, 285, 304

models 103, 138, 170, 174–175sets of 125, 151, 182–183, 299

city marketing 92cluster analysis 108–109, 239cognition 277cognitive mapping 321–322Colorado 253–262competitive advantage 65, 69

competitor analysis 299conation 277confusion 26–28consideration cycle 154constraints theory 171–174

types of 172consumer choice 3, 14consumer psychology see psychology, consumerconsumption 22, 135

stages of 3, 122, 133post-purchase 3, 12, 125–128pre-purchase 3, 122, 133purchase and acquisition 3, 122, 128

systems of 75, 77, 80–90, 164content analysis 69, 324contingency tables 249convenience 56conventions 135–145correlation analysis 80cross-cultural issues 4, 303–316

indirect and direct study methods 305–306cultural activities 57cultural determinants 265–273cultural impacts 35cultural value 55, 59, 61culture, Islamic 52curiosity 26–28

decision making 91, 121–133, 150, 152, 161–167ethics of see ethicsmodels of 155–156, 161–162on-vacation decision making 166pre-trip decision making 162–163, 166stages of 122, 166

decision processes see decision makingdecision support system 93–94, 100demand 123

constraints on 128

Consumer Psych - Chap Index 4/12/03 4:24 pm Page 333

demographic profiles 77desire 193–194destinations 77, 104

characteristics 231, 267choice of see choice of destinationcities 85–88coastal 81–82competitiveness of 286, 288, 301, 304familiarity with 164image of 268–269, 278, 289, 303–316, 320,

328–330life cycle 39, 45, 299loyalty toward see loyaltymarketing 279performance of 285–301secondary 162–163

Differential Emotional Scale 23, 25dissatisfaction 12, 16–17, 23, 26–28disappointment see dissatisfactiondispleasure see dissatisfactiondomestic tourism 51–62, 75–90dominance 22

e-commerce 238economic factors 55economic impacts 35, 41ecotourism 65–72

definition of 66emotional contagion 29emotions 21–30employees see service employeesemployment in tourism 218, 223–225environmental impacts 35, 41environmental information 95environmental learning 65–66, 67–68, 70environmental psychology see psychologyethics 215–225

beliefs 217–218divergence between staff, management and

visitors 219–223measurement of 218values see values

expectations 12, 16, 17, 22, 24, 26, 191, 194experiments 94, 138, 185exploratory behaviour 276exploratory data analysis 246–247expressive factors 227–235

factor analysis 26–28, 41, 54, 61, 124, 127, 197–198,205–210, 219, 231, 311–313

familiarity 234family togetherness 55fear 26–28first-time visitation 103–118, 204–213

Florida 203–213frankness 219–222friendliness 219–222

game theory 92 97–100guests see visitor typesGuttman scale 45–46

happiness 26–28Hawaii 265–273helpfulness 219–222heuristics 96honesty 219–222host communities 256, 259–260, 265hotel consumers 11–18hotel selection 13

image 278information 104–105, 108

format 184processing 12, 17search 149, 153spatial 116

instrumental factors 228–235intention to return see repeat visitationinterpretation 68Internet 93, 237–243involvement 279Ireland 321–330

knowledge 55

learning 122leisure

motivation 58time 76travel 76–77

LISREL see structural equation modellinglocation 15–16logistic regression 239–240loyalty 266, 275–282

latent loyalty 277spurious loyalty 277

Mallorca 285–301market segmentation 11, 18, 90, 238–243,

245–252marketing strategy 90mood 21–22motivation 52–62, 67, 127, 65, 176, 265, 324

Maslow’s hierarchy of 53, 67studies of 52

334 Index


nationality 270nature activities 57Netherlands 106New Zealand 36, 39novelty 24, 276, 278numen seeking 68

operator model 152–158over-commercialization and over-development 298

packaging 181–187perception 128, 269perceptual mapping 313–315personality 170–171, 176pleasure 22Portugal 121–133price 18principal component analysis see factor analysispreference

revealed 138stated 138

psychological attachment 277–278psychology

branches of 2cognitive psychology 2consumer psychology 1

elements of 4in tourism, hospitality and leisure 1–9

context 2dimensions of 2economic psychology 3environmental psychology 2, 4, 22levels of analysis 2social psychology 2

push and pull factors 52, 59–61

quick clustering 79–80

regression analysis 60, 206–207, 210–211, 220relaxation 56religion 58–59repeat visitation 104–118, 204–213, 265–273, 278resident attitudes towards tourism 33–46revealed choice 107risk 12, 15, 17–18, 123, 279route choice 96

sadness 26–28safety 57Saudi Arabia 51–62satisfaction 5, 21–23, 25–26, 30, 189–200, 203–213,

227–235, 266, 319, 327antecedents and consequences of 190–191model of 229

self concept 193, 195–196, 198, 200self-organizing feature maps (SOFM) 247–250sensation seeking 276, 282sequence alignment method 105, 117–118service employees 256–58service encounters 21, 216–225, 256service experience 22service quality 21–30, 39, 190, 192–193, 223–225, 257services marketing triangle 255–256Servperf 25Servqual 24–25simulation 94, 97site selection 137, 144ski resorts 253–262skiers 260–261social distance 33–46, 48–49

internal consistency of 40–41scaling 34, 36, 38–39, 46

social factors 55social impacts 35, 41social interaction 150, 151Spain 305–309stakeholder theory 253, 262stimulation, optimum stimulation level (OSL)

276, 280structural equation modelling 198, 229–233sustainable tourism 253–262

Taiwan 149–158theme parks 104tourism behaviour 76, 266tourism impacts 35, 45tourism marketing 76, 89–90, 92tourist distribution 98–110travel behaviour 230, 307–309travel career ladder 330trust 26–28Tunisia 186Turkey 186, 285–301typology representing networks 248–250

urban tourism 91–100United Kingdom 23–30, 197utility 55, 61, 92, 96–97, 98

values 216–217variety-seeking behaviour 276, 280visiting friends and relatives 84–85visitor types 230–231, 233, 260–261

web see Internet

Index 335


consumer psychology of tourism, hospitality and leisure - consumer... · consumer psychology of...

Documents