guidelines for sensory analysis in food product ...978-1-4615-4447-0/1.pdf · guidelines for...
TRANSCRIPT
Guidelines for Sensory Analysis
• In Food Product Development
and Quality Control
Second Edition
Roland P. Carpenter Sensory Computing Specialist Unilever Research Colworth
United Kingdom
David H. Lyon Head of Consumer and Sensory Sciences
Campden & Chorleywood Food Research Association United Kingdom
Terry A. Hasdell Scientific Services Manager United Biscuits (U.K.) Ltd
United Kingdom
AN ASPEN PUBLICATION® Aspen Publishers, Inc.
Gaithersburg, Maryland 2000
Library of Congress Cataloging-In-Publication Data
Guidelines for sensory analysis in food product development and quality controV[edited by] Roland P. Carpenter, David H. Lyon,
Terry A. Hasdell.-2nd ed. P. em.
Includes bibliographical references (p. ). ISBN 0-8342-1642-6
I. Food industry and trade--Quality control. 2. Food-Sensory evaluation. I. Carpenter, Roland P. ll. Lyon, David H., 1956-
III. Hasdell, Terry A. TP372.5.G85 1999
664'.07-dc21 99-27341
CIP
Copyright ©2000 by Campden & Chorleywood Food Research Association (CCFRA).
Aspen Publishers, Inc., A Wolters Kluwer Company www.aSllenvublishel]£Q!ll All rights reserved.
Aspen Publishers, Inc., grants permission for photocopying for personal or internal use. This consent does not extend to other kinds of copying, such as copying for general distribution,
for advertising or promotional purposes, for creating new collective works, or for resale. For information, address Aspen Publishers, Inc., Permissions Department,
200 Orchard Ridge Drive, Suite 200, Gaithersburg, Maryland 20878.
Orders: (800) 638-8437 Customer Service: (800) 234-1660
About Aspen Publishers· For more than 40 years, Aspen has been a leading professional publisher in a variety of disciplines. Aspen's vast information resources are available in both print and electronic formats. We are committed to providing the highest quality information available in the most appropriate format for our customers. Visit Aspen's Internet site for more information resources, directories, articles, and a searchable version of Aspen's full catalog, including the most recent publications: http://www.aspenpublishers.com
Aspen Publishers, Inc .• The hallmark of quality in publishing Member of the worldwide Wolters KIuwer group.
Editorial Services: Ruth Bloom Library of Congress Catalog Card Number: 99-27341
ISBN: 0-8342-1642-6
I 2 3 4 5
Table of Contents
Contributors .......................................... xi
Preface .............................................. xv
Introduction .......................................... xix
Acknowledgements ..................................... xxvii
Chapter I-What Is Sensory Analysis Used for? .......... . 1.1 Providing Answers to Practical Questions .... . 1.2 Specifications and Quality Control .......... . 1.3 Shelf-Life Studies ....................... . 1.4 Taint Potential .......................... . 1.5 Product Matching ....................... . 1.6 Product Refonnulation ................... . 1. 7 Product Mapping ........................ . 1.8 Product Acceptability .................... .
Chapter 2-The Relationship of Physiology and Psychology to Sensory Analysis .......................... . 2.1 Introduction ............................ . 2.2 What Role Do the Senses Play? ............ .
2.2.1 Sight ............................. . 2.2.2 Smell ............................ . 2.2.3 Taste ............................. .
2.3 How Do the Senses Interact? .............. . 2.3.1 Texture ........................... . 2.3.2 Flavor ............................ .
2.4 Thresholds and Sensitivity
iii
1 1 2 3 5 7 8 9
11
13 13 14 14 17 19 21 21 25 27
iv GUIDELINES FOR SENSORY ANALYSIS
2.5 Individual Differences . . . . . . . . . . . . . . . . . . . . . 28 2.5.1 The Importance ofIndividual
Differences . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.5.2 Gender ............................ 29 2.5.3 Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.5.4 Physiological State . . . . . . . . . . . . . . . . . . . 29 2.5.5 Genetics ........................... 30 2.5.6 Psychology . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.6 Perception-The Link with the Senses .... . . . . 31 2.6.1 How Is Psychology Linked with
Physiology? ........................ 31 2.7 Which Psychological Factors May Affect
Sensory Results? ......................... 31 2.7.1 Biases and Errors .................... 31 2.7.2 Personality and Attitude . . . . . . . . . . . . . . . 34
Chapter 3-How To Use Sensory Analysis To Meet Your Objective ................................. 35 3.1 Formulating the Objective. . . . . . . . . . . . . . . . . . 35
3.1.1 The Key Question To Be Answered ..... 35 3.1.2 Understanding the Client's Need. . . . . . . . 36
3.2 Decisions about Data . . . . . . . . . . . . . . . . . . . . . . 36 3.3 Tests Used To Achieve the Objective. . . . . . . . . 41
3.3.1 Discrimination or Difference Tests ...... 41 3.3.1.1 Paired Comparison (Difference)
Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 3.3.1.2 Duo-Trio Test ................... 42 3.3.1.3 Difference from Control Test ....... 42 3.3.1.4 Triangle Test .................... 43 3.3.1.5 Two-Out-of-Five Test .. . . . . . . . . . . . 44 3.3.1.6 Ranking Test .................... 44 3.3.1. 7 Magnitude Estimation .......... . . . 45
3.3.2 Descriptive Tests .................... 46 3.3.2.1 Consensus Profiling .............. 46 3.3.2.2 Descriptive Profiling . . . . . . . . . . . . . . 47 3.3.2.3 Free-Choice Profiling ............. 48
3.3.3 Acceptance Tests .................... 49 3.3.3.1 Hedonic Rating .................. 50
Table o/Contents v
3.3.3.2 Paired Comparison (Preference) Test . . . . . . . . . . . . . . . . . 51
3.3.3.3 Repeat Paired Comparison (Preference) Test . . . . . . . . . . . . . . . . . 52
3.3.3.4 Multi-Sample Ranking for Preference ...................... 52
3.4 Principles of Questionnaire Design . . . . . . . . . . . 53
Chapter 4-The Products for Sensory Analysis ............ 59 4.1 The Nature of the Products ................. 59
4.1.1 What Are the Products To Be Assessed? .......................... 59
4.1.2 Special Practical Difficulties ........... 62 4.2 The Number of Products ................... 67
4.2.1 How Many Product Types Are There? ... 67 4.2.2 Is It Necessary To Evaluate Every
Product? ........................... 68 4.2.3 How Much Product Is Required? . . . . . . . . 68
4.3 The Assessment of the Products ............. 69 4.3.1 How Is the Product To Be Assessed? 69 4.3.2 What Are the Practical Limits When
Preparing Products for Sensory Analysis? .......................... 69
4.3.3 How Many Products Should Be Assessed in One Sitting? ...................... 70
Chapter 5--Who Are the Right People for Sensory Analysis? ................................. 71 5.1 What Type of Person Is Required? ........... 71
5.1.1 Sensory Analyst or Panel Leader ........ 71 5.1.2 Sensory Assessors ................... 72
5.2 Selecting People for Specific Tasks .......... 75 5.2.1 Quality Control ..................... 75 5.2.2 Taint Tests ......................... 76 5.2.3 Discrimination or Difference Tests ...... 76 5.2.4 Descriptive Tests .................... 76 5.2.5 Acceptance Tests .................... 77
5.3 How Many Assessors? .................... 78
vi GUIDELINES FOR SENSORY ANALYSIS
5.4 How To Select Assessors .................. 80 5.4.1 Selection on Ability .................. 80 5.4.2 Criteria for Selection ................. 82
5.5 How To Train Assessors ................... 82 5.5.1 General Training .................... 82 5.5.2 Training For Specific Tasks. . . . . . . . . . . . 83
5.6 Health and Welfare of Assessors ............ 88 5.6.1 Introduction ........................ 88 5.6.2 Possible Sources of Risk: The
Equipment ......................... 89 5.6.3 Possible Sources of Risk: The Products . . . 89 5.6.4 Possible Sources of Risk: The Test
Environment . . . . . . . . . . . . . . . . . . . . . . . . 90 5.6.5 General Procedures and Protective
Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 5.6.6 Special Procedures within the Test
Protocol ........................... 91 5.6.7 A Practical Example . . . . . . . . . . . . . . . . . . 91
Chapter 6-Experimental Design and Data Analysis ........ 95 6.1 Experimental Design. . . . . . . . . . . . . . . . . . . . . . 95
6.1.1 What Is Experimental Design? ......... 95 6.1.2 What Factors Should Be Taken into
Account in the Experimental Design? ........................... 97
6.2 Data Analysis ........................... 106 6.2.1 What Statistical Methods Are
Available? ......................... 107 6.2.2 How To Decide If the Data Are
Normally Distributed ................. 108 6.2.3 Which Methods Are Appropriate for
Discrimination Tests? ................ 109 6.2.4 Which Methods Are Appropriate for
Descriptive Tests? ................... 109 6.2.5 Which Methods Are Appropriate for
Acceptance Tests? ................... 115 6.2.6 Which Methods Are Appropriate for
Relating Data? ...................... 117
Table of Contents Vll
6.2.7 Hypothesis Testing. . . . . . . . . . . . . . . . . .. 119 6.2.8 Points To Consider in Choosing a
Statistical Package ................... 120
Chapter 7-Reporting and Recording.................... 121 7.1 Style and Content ........................ 121 7.2 Background and Objectives ................ 122 7.3 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 123 7.4 Analysis of Data ......................... 124 7.5 Presentation of Results .................... 124 7.6 Interpretation and Discussion of Results . . . . . .. 126 7.7 Conclusions ............................. 126 7.8 Recommendations ...................... " 126 7.9 Retaining Records ........................ 127
Chapter 8-Putting Sensory Analysis into Practice •..•..... 129 8.1 Resources .............................. 129
8.1.1 Are Suitable Equipment and Facilities Available? ......................... 129
8.1.2 What Are the Requirements of the Test Area? ............................. 129
8.1.3 What Equipment Will Be Needed During Sample Assessment? ................. 131
8.1.4 Will Assessors Be Available for the Test? ........... . . . . . . . . . . . . . . . . . .. 131
8.2 Practical Constraints ...................... 131 8.2.1 Product Constraints: Availability and
Aging ............................. 131 8.2.2 Time Constraints .................... 132 8.2.3 Cost Constraints . . . . . . . . . . . . . . . . . . . .. 132
8.3 Organizing The Sensory Test ............... 133 8.4 Assessor Briefing and Motivation . . . . . . . . . . .. 134
8.4.1 Do the Assessors Know What To Do? . . .. 134 8.4.2 Do Assessors Receive Feedback on the
Test Results? ....................... 134 8.4.3 Should Assessors Receive Any Special
Incentives or Rewards? ............... 135
Vlll GUIDELINES FOR SENSORY ANALYSIS
8.5 Setting Standards for Sensory Panels ......... 135 8.5.1 Product Standards in Discrimination
Testing ............................ 136 8.5.2 Standard Products in Descriptive Tests ... 136 8.5.3 Standard Products in Acceptance Tests. .. 137
8.6 Monitoring Performance of Assessors and the Panel .................................. 137 8.6.1 Monitoring Performance Using Product
Standards .......................... 138 8.6.2 Comparison with Other Assessors . . . . . .. 139 8.6.3 Comparison with Other Panels: Ring
Tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 140
Chapter 9-Case History: Specification and Quality Control ................•................•. 143 9.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 143
Question: What Is the Target Specification and How Can It Be Defined? . . . . . . . . . . . . . . . . . .. 143
9.2 Method and Approaches ................... 143 9.3 Implementation .......................... 144
Chapter lO-Case History: Shelf-Life Studies ...•.......... 145 10.1 Background ............................ 145
Question: How Long Can a Product Be Stored before the Sensory Characteristics Change? How Do Products Change with Storage? ..... 145
10.2 Methods and Approaches ................. 145 10.3 Analysis and Presentation of Results ........ 146
Chapter 11-Case History: Taint Investigation ............. 149 11.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . .. 149 11.2 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 150 11.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 150
Chapter 12-Case History: Taint Prevention ............... 153 12.1 Background ............................ 153 12.2 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 153 12.3 Results ................................ 154
Table o/Contents IX
Chapter 13-Case History: Mapping of Coffee Products ..... 157 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .. 157 13.2 Aims ................................. 157 13.3 Descriptive Analysis of Coffee ............. 158 13.4 Implications for Product Matching!
Mapping .............................. 162 13.5 Consumer Preferences for Coffee ........... 163 13.6 Relating Consumer Preferences to Sensory
Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 166 13.7 Implications for Product Matching!
Mapping .............................. 167
Chapter 14-Case History: Quality Control in Product Batching .................................. 169 14.1 Background ............................ 169 14.2 Methods and Approaches ................. 169 14.3 Recommendations ....................... 170
Chapter IS-Case History: Graphical Methods for Monitoring Profile Panel Performance ................... 171 15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .. 171 15.2 Methods and Results ..................... 171
Plot of Assessor x Product Interaction ....... 172 Plot of Replicate X Product Interaction Plot for an Individual Assessor ................... 173
15.3 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . .. 174
Appendixes 177
Appendix A-Some Useful Tables for Sensory Tests ............................. 177
Appendix B-Glossary of Terms Used in Sensory Analysis. . . . . . . . . . . . . . . . . . . . . . . . . .. 181
Appendix C-Some Useful Contacts ............... 189
Bibliography .......................................... 195
Index 201
Contributors
The publication of these Guidelines/or Sensory Analysis in Food Product Development and Quality Control has been possible only by the dedicated contribution of many people and their willingness to share their knowledge and experience to encourage practical applications of sensory analysis. The editors wish to acknowledge the original contributors to the first edition published in 1992, whose work was the basis for these new Guidelines. However, many more people have contributed many hours of their own time to bring new information into the second edition of the Guidelines and to make them more applicable to the industry today.
Roland Carpenter, Unilever Research Colworth, has masterminded the reorganization and rewriting of key sections of the Guidelines and joins David H. Lyon, Campden & Chorleywood Food Research Association (CCFRA), and Terry A. Hasdell, United Biscuits (U.K.) Ltd., as one of the main editors of this second edition.
The drive toward updating the Guidelines was sustained only with support of the members of the CCFRA Sensory and Consumer Methodology Working Party and in acknowledgement, a list of their names and affiliations is given below. Particular thanks go to Carol Raithatha, Lauren Rogers, Janet Colwill, Alison Atkinson, Margaret Everitt, Stuart Alcock, Jean McEwan, and Zoe Cselik whose extra efforts helped to keep the work on target and allowed the publisher's deadlines to be met.
xi
XU GUIDELINES FOR SENSORY ANALYSIS
SENSORY & CONSUMER METHODOLOGY WORKING PARTY MEMBERSHIP 1998
MrS. Alison Atkinson, Weetabix Ltd., Burton Latimer, Northamptonshire, U.K.
Ms. Julia Bell, Mars Confectionery, Slough, Berkshire, U.K.
Mr. Simon Bellchambers, Kerry Ingredients, Listowel, County Kerry, Ireland
Ms. Stephanie Bettan, Danone Tepral Centre de Recherche, Strasbourg, France
Mr. Roland Carpenter, Unilever Research Colworth, Sharnbrook, Bedfordshire, U.K.
Dr. Pete Channon, Whitbread pIc, Luton, Bedfordshire, U.K.
Ms. Kersten Christensson, Nestle R&D Center: Bjuv AB, Bjuv, Sweden.
Mr. Michael Clements, Guinness Brewing Worldwide Research Centre, Dublin, Ireland
Dr. Janet Colwill, Kerry Ingredients, Bristol, Avon, U.K.
Miss Sandrine Courregelongue, Nestle U.K. Ltd., Croydon, Surrey, U.K.
Dr. Zoe Cselik, United Biscuits (U.K.) Ltd., High Wycombe, Bucking-hamshire, U.K.
Mr. Stuart Davison, Weston Research Laboratories Ltd., Maidenhead, Berkshire, U.K.
Mr. Stephane Durand, Campden & Chorleywood Food Research Associa-tion, Chipping Campden, Gloucestershire, U.K.
Mrs. Jayne Edwards, Premier Brands U.K. Ltd., Wirral, Merseyside, U.K.
Mr. Michael Ellis, Nestle Rowntree, York, North Yorkshire, U.K.
Mrs. Margaret Everitt, Campden & Chorleywood Food Research Association, Chipping Campden, Gloucestershire, U.K.
Mrs. Alison Ferkin, Birds Eye Walls Ltd, Gloucester, Gloucestershire, U.K.
Contributors xiii
Ms. Caroline Giraud, Nestle R&D Centre, Beauvais, France
Ms. Janet Goodman, New Zealand Milk (U.K.) Ltd., Swindon, Wiltshire, U.K.
Mr. Terry A. Hasdell, United Biscuits (U.K.) Ltd., High Wycombe, Buckinghamshire, U.K.
Dr. Paul Hegarty, Bass Brewers Ltd., Burton on Trent, Staffordshire, U.K.
Miss Rachel Hornsby, Reckitt & Colman Products Ltd., Hull, Humberside, U.K.
Ms. Sophie Lefebvre, Nestle U.K. Ltd., York, North Yorkshire, U.K.
Mr. David Lyon, Campden & Chorleywood Food Research Association, Chipping Campden, Gloucestershire, u.K.
Ms. Caroline Martin, United Distillers & Vintners, Menstrie, Clackmannanshire, Scotland, U.K.
Dr. Jean McEwan, Campden & Chorleywood Food Research Association, Chipping Campden, Gloucestershire, U.K.
Mr. David Mortimer, Van den Bergh Foods, Ltd., Purfleet, Essex, U.K.
Mr. Richard Neish, UDV (HP) Ltd., Bishops Stortford, Hertfordshire, U.K.
Ms. Sally Page, Tesco Stores Ltd., Cheshunt, Hertfordshire, U.K.
Miss Ann-Marie Potts, McCormick (U.K.) pIc, Aylesbury, Buckinghamshire, U.K.
Mrs. Mary Quinlan, Tate & Lyle Specialty Sweeteners, Reading, Berkshire, U.K.
Mrs. Carol Raithatha, British Sugar Technical Centre, Norwich, Norfolk, U.K.
Mrs. Lauren Rogers, DuPont Cereals Innovation Centre, Cambridge, Cambridgeshire, U.K.
Ms. Nicole Snijders, SmithKline Beecham Consumer Healthcare, Coleford, Gloucestershire, U.K.
XIV GUIDELINES FOR SENSORY ANALYSIS
Mrs. Kate Sullivan, Crown Cork & Seal Company Inc., Wantage, Oxfordshire, U.K.
Ms. Ellen Taylor, Bush Boake Allen Ltd., Walthamstow, Greater London, u.K.
Miss Susan Tilbury, The Tetley Group Ltd., Greenford, Middlesex, u.K.
Mrs. Jane Walton, Elida Faberge, Leeds, West Yorkshire, u.K.
Ms. Gillian Wright, Sainsbury's Supermarkets Ltd., London, U.K.
Preface
Sensory testing has been in existence ever since man started to use his senses to judge the quality and safety of drinking water and foodstuffs. With the onset of trading, there were several developments that led to more formalized testing, involving professional tasters and grading systems. Many of these grading systems are still in existence today and continue to serve a useful purpose, for example in assessing tea, coffee, and wines. However, there has also been a growing need for methods for well-replicated, objective, unbiased sensory assessment, which can be applied routinely across a wide range of foods. Sensory analysis seeks to satisfy this need.
Sensory analysis is not new to the food industry, but its application as a basic tool in food product development and quality control has not always been given the recognition and acceptance it deserves. This, we believe, is largely due to the lack of understanding about what sensory analysis can offer in product research, development, and marketing and a fear that the discipline is "too scientific" to be practical. To some extent, sensory scientists have perpetuated this fear by failing to recognize the industrial constraints to implementing sensory testing procedures. These Guidelines are an attempt to redress the balance.
Of course, product "tasting" is carried out in every food company. It may be the morning tasting session by the managing director; comparisons of competitor products by the marketing staff; quality assurance tasting by a product "expert"; evaluation of new recipes from the product development kitchen; or on-line checking during production. Most relevant, though, is that the people responsible for the tasting session should know
xv
xvi GUIDELINES FOR SENSORY ANALYSIS
why the work is being done and fully realize that if it is not done well, the results and conclusions drawn and their implications are likely to be misleading. If these Guidelinesfor Sensory Analysis in Food Product Development and Quality Control (Guidelines) help to persuade those who organize and participate in food tasting in companies to review and reassess what they are doing, and why, we believe our efforts will have been worthwhile.
All those who have contributed to these Guidelines are practical sensory analysts with a wide range of experience in using sensory analysis in industrial situations. Our numbers include sensory analysts working in major European food companies and sensory scientists from the Campden & Chorleywood Food Research Association. We are all members of the Sensory and Consumer Methodology Working Party (fonnerly the Sensory Evaluation Working Group), an active forum for the discussion of new research and the exchange of practical experiences in the use of sensory analysis. The recommendations we offer are all based on many years of practical experience. We continue to make use of these Guidelines widely throughout our own companies in order to help our colleagues attain a greater understanding of the usefulness and relevance of sensory analysis techniques when applied in a controlled and appropriate manner.
With the wider industrial interest in mind, these Guidelines make reference to standard generic sensory methods (e.g., ISO methods) that can generally be applied to all foodstuffs across all food and drink industries. Some industries have published additional sensory methods that are specific to their given type of food or drink, e.g., the beer tasting methods published by the Institute of Brewing (Institute of Brewing, 1995). These industry methods can provide useful additional infonnation on specific flavor descriptors for given product types, reference materials for training, the methods of handling specific food or drink products, etc. It is always worthwhile to check whether any relevant industry sensory methods have been published before setting up new sensory analysis methodology.
When first considering the need for a Guidelines book, we were aware of no other publication specifically designed to help not only those who are carrying out sensory testing but also those within the industry who are making use of sensory analysis results. It is our opinion that both groups would benefit from a clearer explanation of the role of sensory analysis in product development and quality control. We had as our objective, therefore, the need to prepare a book that would promote the use of appropriate
Preface xvii
sensory analysis techniques within the food industry and would encourage the greater application of sensory analysis as an integral part of industrial product development and quality control.
The main part of these Guidelines has been written to follow a logical sequence of questions that might be asked before embarking on sensory analysis. We have tried to make the layout user-friendly, using a question and answer format that can be dipped into as and when guidance is required on particular issues. We have included brief descriptions of some of the most widely used sensory tests and provided readers with details of where they might find more information about particular tests, if we believe they are adequately documented elsewhere. References to suitable and helpful literature are given, together with a glossary of terms. We have drawn on our experiences to provide case histories and examples of when sensory tests have proved useful in particular situations.
Since the first edition of these Guidelines was published, we have seen major developments in standardization and objectivity. International and British standards continue to prescribe approved methods for designing sensory facilities, recruiting panels, and running sensory tests. Companies are increasingly recruiting external assessors to specialize in undertaking sensory work. There has been an increase in the use of continuous line scales as the basis for sensory measurement, and research continues into new sensory and statistical methods. In this context of developing professionalism, the establishment of a formally recognized sensory analysis qualification now seems long overdue.
We hope to illustrate that good sensory practice is not just a question of the right resources, but also requires scientific discipline and focus. Whatever the size of the company, whatever the standard of the sensory facilities, if the approach takes into account the ideas we are offering in these Guidelines, we are sure that it will result in greater confidence in sensory testing, greater reliability in the application of the results of sensory analysis, and a significant commercial benefit.
Introduction
Perhaps you believe that sensory analysis may have something to offer your company, either as an internal function or as a service bought in on a contract basis. Perhaps you have been asked to set up your own sensory testing service and are looking not only for guidelines, but also for arguments to justify and sell your involvement in commercial projects. Perhaps you are studying sensory analysis and are looking to see how it is best applied in the food industry. Whoever you are and whatever your purpose for delving into this book, here is the chance to explore a few fundamental questions that may be going through your mind.
• What is sensory analysis and what does it tell us? • What questions does sensory analysis answer? • Why should we use sensory analysis? • How does sensory analysis help? • What are the benefits of using sensory analysis? • How do we get started?
WHAT IS SENSORY ANALYSIS AND WHAT DOES IT TELL US?
Sensory analysis is the identification, scientific measurement, analysis and interpretation of the properties (attributes) of a product as they are perceived through the five senses of sight, smell, taste, touch, and hearing.
XIX
xx GUIDELINES FOR SENSORY ANALYSIS
Sensory analysis has to do with measurement. However, there is nothing in this definition to indicate whether the measurement is qualitative (X is sweeter than Y) or quantitative (on a sweetness scale from 0 to 100, Y receives a rating of 45 and X a rating of70). Nor does the definition stipulate whether trained assessors or untrained consumers are making the assessments, although this is just one of the important decisions that must be made when planning a sensory study. A further and more fundamental decision concerns the basis of the question or assessment, whether it is an objective analytical quality such as flavor intensity or firmness, or a subjective value judgment such as liking, preference, or acceptability. The classical view is that sensory analysis, with its roots in science and the laboratory, is only concerned with objective sensory questions, and that subjective quality is based in the separate sphere of consumer research, led by marketing departments or social scientists.
Sensory analysis and consumer research are both concerned with obtaining answers to particular questions relating to the perception of product quality. What is most important, particularly in an industrial context, is that the right questions are posed, in the right way, to the right audience, at the right time. One of the purposes ofthese Guidelines is to help to identify the right way to handle each investigation.
There is a growing trend to integrate what were traditionally seen as separate sensory analysis and consumer research methods into a single array of complementary product tests focusing on sensory questions, each tailored to a particular situation and each contributing a distinctive element of information. There are three main stages. In markets driven by consumer choice, it is customers' expectations and ideals that determine new product targets, and these ideals must be identified through preliminary qualitative consumer research at stage one. The second stage is the efficient design and development of those product targets, which relies increasingly on well-structured sensory analysis studies. The final stage is a return to classical consumer research in order to conduct a quantitative test of product preference and acceptability.
WHAT QUESTIONS DOES SENSORY ANALYSIS ANSWER?
Sensory analysis answers questions of quality under three main headings--discrimination, description, and preference.
Introduction xxi
Discrimination
These questions aim to find out whether or not a difference exists be-tween two or more products.
• Is this product the same as that? • Is this one different? • How great is the difference? • Would people notice the difference? • Would people detect this taint? • How many would discriminate/detect this difference?
Description
These questions aim to describe and measure any differences that are found to exist between products.
• What does this product taste like? • What are its perceived sensory characteristics/attributes? • How does a change in process/formulationipackaging/storage con
ditions affect the sensory quality of this product? • How do these products differ from one another in sensory quality? • For which sensory attributes are the differences most marked?
Preference or Hedonics
These questions aim to identify liking or acceptability.
• Do you like this product? • How much do you like this product? • Is it acceptable? • Is it as good as this other product? • Is this product better than that one? • Do you prefer this one? • What do you most like about this product? • Which other features/attributes do you like?
WHY SHOULD WE USE SENSORY ANALYSIS?
Product tasting is undertaken partly to communicate and partly to provide an input for decision making. We should clearly distinguish these two
xxu GUIDELINES FOR SENSORY ANALYSIS
objectives and try to ensure that the structured framework of sensory analysis is employed in all those product tasting situations on which major decisions depend.
In all food companies, whether manufacturing or retail, the tasting of products is practically a daily occurrence. There are many reasons for looking at products: to maintain awareness of your own company or competitor products; to promote your company products to potential customers; to seek a customer's approval; to demonstrate the latest new product range to the sales team; to keep the project team or management up to date on process or product development progress; to decide whether to proceed with formulation or process changes; to check that product quality matches a target or meets a specification. All except the last two purposes are largely exercises in communications, maintaining awareness, and promoting products to secure management or customer approval. The last two examples illustrate where sensory information is required in order to make decisions about product quality, whether to follow or change a line of research or development, and whether to release product for distribution.
The informal tasting environment may provide good opportunities for ad hoc opinions, observations, and comments. However, this informality, lack of structure, and methodological discipline makes it quite the wrong environment for collecting reliable sensory information to support major decisions. Instead, we should adopt the properly structured framework of sensory analysis, to provide high-quality information on which those critical business decisions can then be more soundly based.
Of course, the quality of the sensory information will affect the quality of the business decisions. That is not to say that sensory analysis should take over any decision making about quality, or that it should seek to replace tasting for communication, awareness, or promotional reasons. It is simply a case of making sure that sensory analysis procedures are used whenever it is appropriate to do so; namely, whenever there is a need to make significant business decisions that are highly dependent upon assessments of product quality.
So, what are the pitfalls associated with obtaining definitive sensory information from informal tasting sessions? Commonly, they fall into the following six categories: the wrong objective, the wrong participants, the wrong questions, biased judgments, lack of scientific control, and the wrong environment.
Introduction xxiii
1. Setting the Right Objective
In the absence of fonnal guidelines, proper assessor training, and clearly defined references, it is easier to say how much you like a product, rather than to analyze its quality, so subjective judgments are often made, inappropriately. Of course, it is legitimate to ask about product liking if you want to know whether to improve product quality, but this infonnation is usually oflittle value if you want to know how to improve product quality.
2. Asking the Right People
At infonnal tastings, the people who are giving their opinions on product quality or liking are seldom typical of the product user or customer and are usually too few in number to be considered a representative sample. Nevertheless, this approach is often justified by claims that companies know their customers, even if there is little or no consumer research available to describe in sensory tenns what the customer does expect from the company's product. At the tasting session, is the question asked and answered in tenns of what the customer would like? Often, the opinions given are personal ones.
The issue of using the right people to make the assessment also applies when addressing the more objective and analytical sensory questions. Are the people assessing the products sufficiently sensitive to the relevant food stimuli? Are they able to discriminate? Are they trained or otherwise able to be analytical in approach and articulate about their perceptions?
3. Asking the Right Questions
Infonnal tasting sessions are not usually planned to answer clear objectives with structured questionnaires, so the infonnation that is offered is sometimes not fully recorded and sometimes relies only on an individual assessor's written comments. In the absence of defined and common terminology, infonnation is patchy, and any attempt at interpretation of comments is risky. Without a structured framework, particular aspects may be overlooked by some assessors, and thus the product picture is incomplete.
4. Eliminating Bias
Judgments may be biased; for example, if an assessor has preconceptions of a particular product, particularly if the product is not presented
xxiv GUIDELINES FOR SENSORY ANALYSIS
under code ("blind"). Bias may also arise if there is some vested interest in the outcome of the assessment, ifthere is discussion during the tasting, or if the opinion of a senior or dominant person has already been stated. The order of tasting of products is another major source of bias in these situations.
5. Achieving Scientific Control
Control over the sample size, composition, temperature, and timing is important. If participants are free to attend in their own time and help themselves to portions of product in any order, any differences in conclusions may be attributable to differences in portion size, composition, or serving temperature, rather than differences between products.
6. Having the Right Environment
The product presentation and tasting environment also influences the outcome of a tasting session. The ability of people to judge food is affected by distractions like noise, including talking, intrusive odors, excessive heat or cold, or other discomfort, so it is important to eliminate these factors if at all possible.
HOW DOES SENSORY ANALYSIS HELP?
Sensory analysis provides the scientific discipline to tackle many of the difficulties identified in the previous section by offering scientific approaches to obtaining complete and appropriate information about product sensory quality. Sensory analysis is all about ways of removing or controlling sources of unwanted error, through control of the environment and sampling, through good experimental design, and by selecting the most appropriate human 'instruments' to make the measurements. It is about asking the right questions and taking into account what we know about human perception.
WHAT ARE THE BENEFITS OF USING SENSORY ANALYSIS?
Sensory analysis is a scientific discipline that is concerned with providing answers to questions about product quality, questions concerning new
Introduction xxv
and existing in-house and competitor products, questions that are posed frequently by managing directors, marketing managers, factory managers, and quality assurance managers, as well as technical, development, research, and production staff. They are questions that are crucially important to the survival of a food business. The right answers to these questions will lead to more informed and more effective business decisions, which in turn should generate business success in the marketplace.
We should also consider the implications of getting the sensory quality answers wrong! In the worst case, the product will fail immediately or progressively lose market share because it fails to deliver the customers' expectations of quality. Poor product information could lead to miscommunication, wrong decisions, wasted consumer research, and much wasted time, effort, and material costs in development and production.
Sensory analysis can and should play an integral role in the management of product quality in the food industry . Its benefits will be measured in terms of shorter lead times for product development and in the accumulation of increased understanding of product behavior. The established use of sensory analysis methodology will also convey an impression of professionalism, which will benefit the company in its dealings with its customers.
HOW DO WE GET STARTED?
First of all, the venture must be visibly supported by senior management and someone must be made responsible for sensory analysis in the company. The next thing will be to identify suitable facilities. Ideally, this would include a food preparation area, specially designed sensory testing booths, a briefing and discussion area, and computing facilities. However, existing facilities could provide a feasible starting point provided that products could be prepared and served hygienically in standardized conditions, and that assessors could undertake their tasks independently and quietly in a room that is free from extraneous odors, noise, and distractions and is evenly illuminated.
Sensory assessors will then need to be recruited. For tasks requiring little or no training, it will be sufficient to draw up a register of volunteers willing to take part. For tests requiring significant training, potential assessors will need to be screened for sensory ability.
The designated sensory analyst will need to become familiar with all the basic sensory tests, procedures, and data analysis methods, with guidance
XXVI GUIDELINES FOR SENSORY ANALYSIS
from this book and its references. It is helpful to have a statistician on hand for advice on statistical design and analysis because many of the computerbased sensory data analysis packages currently available still provide only limited statistical guidance.
Finally, put it all into practice. We hope these Guidelines help, but there are some things to say at the outset. Essentially, keep it simple and follow these golden rules:
1. Be clear on test objectives. Think about the possible outcome of a test and identify in advance the appropriate actions for each outcome (action standards).
2. Do not make things too complicated for your assessors. In considering your objectives, break complex questions into a series of simple ones. Do one test at a time to avoid bias and confusion.
3. Recruit the right assessors for your test. 4. Ensure that your assessors are clear on their task and fully trained. 5. Design your test to minimize error and bias from assessors and prod
ucts. 6. Create the best environment you can for the analysis. 7. Minimize sources of error from product origin, preparation, and
serving. Carefully control timings, temperatures, portion sizes, and ingredient concentrations, e.g., salt, sugar, milk.
8. Do not try to assess too many samples at once. Break them into small sets and provide rest periods in between.
9. Know how to analyze and present your results for best effect.
We wish you every success!
Acknowledgments
SECTION 5.6.7:
The editors wish to thank the Institute of Brewing Analysis Committee for permission to reproduce their General Policy Statement on Health and Safety Aspects.
CASE HISTORY 13:
The editors thank the European Sensory Network for allowing this case history to be used. A book giving full details of this coffee study has been published (ESN, 1996).
XXVll