integrating deductive

8/3/2019 Integrating Deductive

http://slidepdf.com/reader/full/integrating-deductive 1/8

103

Introduction

Opinions are divided am ongst researchers as

to what constitutes legitimate inqu iry and

warrant able knowledge in specific situations.

Indeed there appear to be two diametrically

opposing views. On the one hand there is the“experimentalist”, “hypothetico-deductive”

or “positivist” and on the other the “natu ralis-

tic”, “contextual” or “interpretative”

(Henwood and P idgeon, 1993, p. 15). One

starts from the need to test theory and th e

other to develop theory. T his creates a dilem-

ma for the researcher faced with a research

question where theory exists but m ay not be

appropr iate in t heir particular circumstances.

T his is the problem we faced. We were inter-

ested in exploring how entrepreneurs used

trust to mitigate customer-perceived risk in

start-up situations. Whilst th ere is an exten-

sive literatu re on the developm ent of tru st

(Mayer et al., 1995: McAllister, 1995) which

allowed the form ulation of a deductive

research design, it had not been applied in this

situation and ou r experience in th e field sug-

gested that it m ay not be entirely appropriate.

We needed to incorporate an inductive

approach into the design.

T his paper draws upon a study of the ways

in which entrepreneur s use trust to mediate

custom er perceived risk at the star t of a ven-

ture in ord er to show how researchers can

combine elements of both app roaches in an

epistemologically consistent way.

The hypothetico-deductive approach

T he p urist hypothetico-deductive perspective

“... emphasises un iversal laws of cause and

effect on an explanatory framework which

assumes a realist ontology; that is that realityconsists of a world of objectively defined

facts” (H enwood and Pidgeon, 1993, p. 15).

In th e deductivist trad ition the researcher

starts “... with an abstract , logical relationship

among concepts then m ove(s) towards con-

crete empirical evidence”, (N euman , 1997,

p. 46). T hus in deductivist research there is a

well-established role for existing theor y since

it informs the d evelopment of hypotheses, the

choice of variables, and the resultant mea-

sures which researchers intend to u se. Withinthis paradigm the scientist formulates a par-

ticular theoretical framework and then sets

about testing it. In an example of this

approach which was relevant to our research,

Qualitative Market Research: An International Journal

Volume 2 · Number 2 · 1999 · pp. 103–110

© MCB University Press · ISSN 1352-2752

Integrating deductiveand inductiveapproaches in a study

of new ventures andcustomer perceived risk

Haider Ali and

Sue Birley

The authorsHaider Ali is a Lecturer in Marketing at Cranfield

University, Silsoe, UK.

Sue Birley is a Professor at The Management School,

Imperial College, London, UK.

Keywords

Consumer behaviour, Consumer’s risk, Entrepreneurs,

Qualitative techniques, Risk management, Start-ups

Abstract

Draws upon a study of t he ways in which entrepreneurs

use trust to mediate customer perceived risk at the start of

a venture in order to show how researchers can combine

elements of both approaches in an epistemologically

consistent way. Specifically, researchers seeking to use an

inductivist/quali tative approach can start wi th an a priori

specification of constructs, perhaps in the form of a model.

One of the ways in which this can help researchers is to

identify where they should look in order to find the

phenomena of int erest to them. We argue that the differ-

ence between inductivist and deductivist research is how

they draw upon existing research: in inductivist research

theory can be used where it is composed of constructswhil e theory represented in the form of variables is more

appropriate in hypothetico-deductive research.



Moorman et al. (1993) study the factors that

determ ine users’ trust in their researchers.

T heir theoretical framework shows that vari-

ous antecedents influence “user trust” in the

researcher and, in turn, th is influences the

utilisation of market research information.

For example “perceived researcher interp er-

sonal characteristics” are an antecedent to

trust; one of the component s of these is the

“perceived expertise” of researchers. Their

theory hypothesises a relationship between

user tru st in the researcher and researcher

expert ise. The basis for th e hypothesis lies in

previous work by Crosby et al. (1990) . Since

the researchers have specific measures for

expertise they are able to t est whether the

hypothesised relationship actually exists.

Qu ant itative or “logical positivist/quan tita-tive” methods (Deshpande, 1983) for data

gathering and analysis are comm only associat-

ed with such approaches.

T he value of such an approach is that

researchers are able to make use of previous

researchers work. H owever, its limitat ion is

that it is only possible to test whether or not ,

or to what extent, the hypothesised relation-

ships exist. T his approach does not help the

researcher to identify what other unan tici-

pated factors may exist such as, for example,

contingent variables or new constru cts. More-

over the researcher can lose the richness of

data which respondent s can provide as a

conversation develops in a more unstructu red

setting (Mintzberg, 1979; Pettigrew, 1988).

T he qu alitative approach does not suffer from

these shor tcomings.

The qualitative approach

Van M aanen (1979) says that the term “quali-

tative” has no precise meaning, that it is anum brella term which covers a variety of tech-

niques, “... which seek to describe, decode,

translate and otherwise come to terms with

the meaning, not the frequency of certain

more or less naturally occurring phenom ena

in the social world” (Van M aanen, 1979,

p. 520) (our italics). In order to gain that

meaning qualitative m ethods emphasise,

“... the representation of reality through the

eyes of participants” (H enwood and Pidgeon,

1993, p. 16). T he focus is on the respondentand it is their reflections and opinions that

should guide the research, so that “... a

qualitative researcher begins with a research

question and little else but b egin(s) with

deta iled observations of the world and

move(s) towards more abstract generalisa-

tions and ideas” (Neum an, 1997, p. 334).

T his distinction between qualitative and

inductivist research on the one hand and

quan titative and deductivist research on the

other is also made by Deshpande (1983). H e

draws on Reichardt and C ook (1979) to

conclud e that an area of different iation

between the qualitative and quant itative

parad igms is that in the qualitative paradigm

research is “grou nded, discovery oriented,

exploratory, expansionist, descriptive,

inductive” whilst the quantitative paradigm

research is “ungrounded, verification-

oriented, confirmator y, reductionist, inferen-

tial, hypothetico-deductive” (D eshpande,

1983).We would argue that this is not a helpful

dichotomy and that using theory to drive

qualitative research, thus introducing a degree

of deductivism to the d ata gathering process,

does not rule out our ability to describe and

explore per se although it may reduce the

extent to which we can explore (see, for

example, Fetterman and Pitman, 1986).

It is how theory is used that matters

Whereas the hypothetico-deductive approach

starts with theory expressed in the form of

hypotheses, which are then tested , qualitative

research avoids this, in order to prematurely

close off possible areas of enquiry (Bryman ,

1988). I f theory does play a role, it is later in

the research p rocess: “... the belief (with

which qualitative research is more commonly

associated) that theoretical reflection ought to

be delayed until a later stage in the research

process” (Bryman, 1988, p. 91).

If it were widely recognised that the ab overecommendation could be followed prod uc-

tively, then there would be no need for the

discussion in this paper. H owever, it has been

recognised that moving away from such a

“purist” approach can have benefits. Indeed,

Eisenhard t (1989, p. 536) notes that

researchers can benefit from an: “... a priori

specification of constructs” which “ ... can

help shape th e initial design of theory building

research”. In fact starting with a completely

clean slate has been argued to be very rare. AsBryman (1988, p.73) notes “Ethnographers

rarely adopt a stance of being ‘sponges’

whereby they simply absorb the subject’s

interpretations”. T his implicit recognition

104

Integrating deductive and inductive approaches

Haider Ali and Sue Birley


Volume 2 · Number 2 · 1999 · 103–110



that p ure indu ctivism may be d ifficult to

pract ice has been made explicitly “although

the qualitative and quan titative approaches

are polar opposites: it should be kept in m ind

that individual researchers in all areas, includ-

ing marketing, fall somewhere along the

continuum between the two extremes”

(Deshpande, 1983, p. 104). T here is, howev-

er, some reticence about going much further.

Eisenhard t (1989, p. 536) cautions

researchers that they should avoid: “thinking

abou t specific relationships between variables

and theor ies”. Neverth eless, it is possible for

the induct ivist researcher to acknowledge

both their interest in specific constructs, and

also their und erstanding of the relationships

between them. What Eisenhard t cautions

against is specifying relationships betweenvariables.

Clearly, this distinction between constructs

and variables is important. Bacharach (1989)

citing Kaplan (1964, p. 55 ) says, “constructs

may be defined as terms which though not

observable either d irectly or ind irectly may be

applied or defined as the basis of the observ-

ables”. Bacharach (1989) a lso cites Schwab

(1980) for the definition of a variable as “an

observable ent ity which is capable of assum -

ing two or more values”. So, for example,

“performance” is a construct for which

“sales” or “retur n on investment” is the

variable.

Developing models using constructs

Up to now researchers have distinguished

between inductivist and hypothetico-

deductive research on the basis of the

presence or absence of theory. We would

argue that there can be a middle ground – one

where existing theor y is used but is presentedin the form of constructs rather th an vari-

ables. This would be synergistic with the

qualitative approach to research, since the

whole tenor of a data gather ing exercise which

is premissed on constructs rather than vari-

ables can be more fluid and adaptive to the

needs of the respondent . T his enables the

researcher to “d iscover” issues or effects

which they may not have had in mind when

the investigation began. So, for example,

asking a respondent about the performance of their firm leaves it open for a discussion of any

one of a nu mber of variables (e.g. sales,

market share, profitab ility). In deed, t aking

this approach assumes that the responden t

would ident ify and focus on the variables

most importan t to them. In contrast, if the

researcher specifically asks about profits, th is

would close off some potential areas of

enquiry. Consequently, we would argue that

the qu alitative researcher can use mod els to

guide their investigations but that they should

be composed of constructs rather than

variables.

T here are two add itional advantages in

using models composed of constructs. First,

since the qualitative researcher is often

advised to deal only with general them es

rather than specific questions, this means that

different respondents may well discuss differ-

ent variables. In such a situation the a p riori

specification of constr ucts provides a useful

means of making sense of the disparate infor-mation provided by various respondent s.

Moreover, since in the subsequent analysis

the researcher is challenged to ident ify the

links between variables and con structs, this

approach allows the responden ts to help to

explain what the relationships are in their

particular circumstances. Fur therm ore they

can help to distinguish between different

constructs.

Second, con structs provide a focus for

research but un like variables they are inher-

ently more general and as such leave open the

scope for generating un intended findings.

After all, in reality there is a clear pressure on

researchers to generate find ings about a

specific topic – to focus their research. Id enti-

fying specific constru cts with which to work is

an aid to this. T his does not, however, pre-

clude the opportu nity to find new constructs

which are relevant to the research question.

Ind eed, this flexibility is a real strength of the

approach we espouse.

Table I shows the differences between thepurist inductive and deductive approaches

and our view as to a combinat ion of the two.

How existing theory can be used inqualitative research: a case study

Having argued that qu alitative researchers can

use models composed of constructs the ques-

tion is how in practical research they can be

used. T he following discussion addresses

some practical issues as to how theory can beused in qualitative research. We consider in

turn the help theory can provide in choosing a

context for the research, the choice of ques-

tions to be asked of respondents and finally

105




Volume 2 · Number 2 · 1999 · 103–110



the tabulation of interview data. To illustrate

our argument, we demon strate the approach

used in our exploration of the u se of trust in

mitigating customer-perceived risk in new

ventures. It is an apposite time in the develop-

men t of this part icular field which is still

relatively new. For example, Sexton (1987,

p. 402) notes that “In 1980, the field was

beginning to emerge” and “research meth od-

ology has also advanced significant ly, bu t th e

application of more advanced techn iques is

still not a standard of the research”. As Smilor

(1987, p. xii) notes, we continue to be con-

cerned to und erstand “what he or she (the

entrepreneur) does”.

D eveloping the mo del

In designing their stud y, researchers may start

with an interest in a par ticular construct or

phenom enon. In our case, we began with a

curiosity as to how entrepreneurs used “tr ust”

to generate sales opportu nities in the creation

of a new ventu re. Whilst there was an exten-

sive exploration, an d discussion, of this

construct in the literature, to our knowledge,

it had rarely been explored in the

entrepreneur ial context. N evertheless, the

literature distinguished between the different

bases that may lead to on e individual trusting

another. It may be because of:

• T heir particular personal characteristics

(characteristic based trust). For example,

they may be mature and carry gravitas

(Rotter, 1980).

• T heir previous interactions (process based

tru st). For example, they may have worked

together as customer and supplier in a

previous employment (Good , 1988).

• T he rule of law (institutional based trust).

For example, the ind ividual is a qualified

doctor or works within a recognised and

respectable organisation (N eu, 1991).

T his existing knowledge was impor tant. It

allowed us to develop the model shown in

Figure 1 and helped us to identify the imper-

sonal and personal bases of trust, the m echa-nisms individuals use to find out whether

someone else is trustworthy, and the specific

reasons why customers may think an entrepre-

neur is trustworthy. As well as helping us to

106




Volume 2 · Number 2 · 1999 · 103–110

Table I The integrated approach compared to purist versions of the deductive and inductive approaches

Stage Purist deductive Purist inductive Integrated approach

1. Develop theoret ical Area of enquiry identifi ed – Develop theoretical

framework but no theoretical framework framework based on constructs

2. Variables identified for Respondents identify Some variables identified

relevant constructs const ruct s and explain t he for relevant construct s –relationship between them others can be identified

by respondents

3. Instrument development Broad themes for discussion Researcher converts the a

identified priori theoretical framework

into atheoretical questions

4. Respondents give answers Respondents discuss general Respondents discuss the

to specific questions themes of interest seemingly general questions

and identify constructs which

are meaningful t o them and

explain the realtionships

between the constructs5. Answers analysed in terms Researcher develops theory Respondent data analysed

of prior theoretical framework on a purely inductive basis according to existing theory.

OR theory is developed on an

inductive basis – without

regard to the existing theory.

6. Outcome Outcome Outcome

Theory tested according to Theory developed Either

whether hypotheses are Existing theory is adapted

accepted or rejected Or

Alternative theoretical

framework is presented



“recognise” the relevant evidence when we

came across it, the model was also useful in so

far as it helped us to con textualise our find ings

in term s of existing research. Since the model is

composed of constru cts it is holistic, in the

sense that although it is based on trust between

individuals, it can also be used to categorise the

use of elements of the marketing mix. For

example, usage of warran ties can be

categorised as using institut ional based trust.

On the other hand, the usage of high prices asan extrinsic cue for quality would be cate-

gorised in the same group as reputation, since

both are methods customers use to determine

whether a m arketer is trustworthy.

Selecting the sample

Our study was concerned with the p rocess

used by the ent repreneur in creating their

venture and an examination of the theor y also

helped us to constru ct a sample which allowed

us to captu re the likely diversity of process-

based tru st. In order to achieve this, we used

the following criteria:

• It was important that we included some

people who had had no previous interac-

tion with their target customers, and where

there would be an absence of process based

trust. T his would give us the opportunity

to gather evidence of how entrepreneurs

107




Volume 2 · Number 2 · 1999 · 103–110

Bases of Trust

personal & impersonal reasons whythe customer should trust the marketer

Institutional

because of the rule of lawe.g. warranties

(can lead to emotional trust)

1st Level of distinction:between the state and the individual

2nd Level

of distinction:between theindividual’s

characteristicsand theirbehaviour

Characteristic - Based

because of who they are(can lead to emotional trust)

Process- Based

because of how they behave(can lead to cognitiveand emotional trust)

Reasons why customersthink entrepreneurs are

trustworthy

Ability

Integrity

Benevolence

Customers can find out

whether entrepreneursare trustworthy t hrough

the following:

Reputation

Social learning

Self-disclosure

Demonstration of trust builds trust

Behavioural

Enactment

customer overcomesperceived risk and

enters into exchange –demonstrates trust

in marketer

Figure 1 Model demonstrating bases of t rust based on existing knowledge



make use of tru st where ostensibly none

ought to exist.

• In order to provide for a broad range of

cases, diversity of indu stry was important

(Van de Ven, 1991) .

• In order to screen out “non-serious” busi-

nesses or those with a very limited scope of

operations, we limited our search to those

businesses in the process of raising ventu re

or development capital.

• Since “new venture creations is a continu-

ous process rather than a discrete event”

(Birley, 1984, p. 65) we sought bu sinesses

less than five years old.

For ty entrepeneurs were interviewed from a

range of businesses that included , for exam-

ple, floatation chambers, rowing machines,

soy sauce manufacturing, dental practice, and

credit card software.

T his deliberate choice of sample would not

prejudice the inductive nature of the enquiry;

indeed Yin (1984) recommend s that cases in

qualitative research can be selected either

because similar results are pred icted (literal

replication) or to generate different results for

predictable reasons (theoretical replication).

Existing theor y was also useful in h ighlight-

ing the link between risk and tru st. T his was

an impor tant notion because it provided us

with an additional means both of finding tru st

and also of recognising it. T hus we could

search for the use of tru st in situations where

the entrepreneurs perceived some form of

risk. Clearly, the building of an entirely new

custom er base fulfilled this criterion since the

conceptual relationship had a correlate in th e

practical notion of “customer perceived risk”,

which m arketers can t ry and overcome using

the “elements of the marketing mix” (EM M).

T he literature suggests that the effectiveness

of strategic choices of the m arketing mix may

be due to their tru st-bearing properties. So,

for example, Wiener and Mowen (1986) have

argued that an important trait d etermining

the success of salespeople is how trustwor thy

customers perceive them to be. Pricing can be

used as a means of influencing customer

predictions of produ ct performance. Cus-

tomers may use high price as an extrinsic cue

for inferring produ ct quality (Bearden an d

Shimp, 1982) , i.e. they may demonstrate

greater tr ust in a product if it has a relatively

high price.

Questions sho uld be atheoretical

H aving used theory to guide our choice of

respondents and also the respondents’ activi-

ties on which we wanted to focus, the final

issue was how we should question them.

Whilst the literature did p rovide oppor tun ities

to follow a deductive path and construct some

form of survey instru ment, we rejected th is.

Qu ite simply, we were not convinced that

existing theory was sufficiently robu st to

capture the par ticular complexities of trust

and customer perceived risk in th e entrepre-

neurial situation. T herefore, we continued

down th e qualitative path.

Our aim was to find ou t how entrepreneurs

manage customer perceived risk. Clearly one

mean s of doing this would have been to ask

them directly how this was done. H owever,

that would have run counter to th e need to

collect data in th e interviewee’s own term s.

Naturally it would also have presupposed

their interest in the subject and their linking

trust with risk. Our task was to develop a

means of questioning the respondent in such a

way that th ey would lead us to risk and trust

but in such a way that there should be no

tautological guarantee of this. In order to

maintain objectivity, we had to rely on respon-

dents taking us to trust and r isk through th eirown volition. So althou gh we knew the exis-

tence of the theoretical link between risk, tru st

and entrepreneurs’ marketing activities, in

order to maintain objectivity it was impor tant

that the link was not d isclosed to them.

T he questions we asked were what ele-

ments of the marketing mix they used and

why they used th em. So, for example, we

might say “H ow did you go about getting your

first ord ers?” followed by “Why did you do it

that way?” T he first question requ ired only adescript ive answer, which had no theoretical

implications, but t he second qu estion leaves it

open to the respond ent to say, in their own

terms, that it was customer perceived r isk

which mot ivated usage of the specific EM M

and that it was because of the credibility or

trust associated with the EM M that they felt

that they would be effective. For example a

clothing manufacturer said that they had used

a sales agent because of the credibility he

would br ing when making sales to retailers.

Similarly a fitness machine man ufacturer u sed

retailers rather then sell direct, part ly because

customers believe that the former can be

relied upon to give independent ad vice.

108




Volume 2 · Number 2 · 1999 · 103–110



of their characteristics, may possibly lead to

the constructs the researcher is interested in.

Finally, as previous commentators on

method ology have remarked (Bryman, 1988),

very few researchers star t off with a “clean

slate”, and very few are able to funct ion as

sponges – merely gathering up everything that

is told to them. H aving taken for granted th at

most researchers und ertake some form of

literature review, and also having taken for

granted that researchers develop interest in

some issues or constru cts, we have sought to

show why an a priori interest should not

reduce the quality of research or indeed be

“glossed” over in em barrassment.

References

Ali, H.A. and Birley, S. (1998), “ The role of t rust in t he

marketing activities of entrepreneurs establishing

new ventures” , Journal of Marketing Management,

Vol. 14, pp. 749-63 .

Bacharach, S.B. (1989), “ Organisational t heories: some

criteria for evaluation” , Academy of Management

Review, Vol.14 No. 4, pp. 496-515.

Bearden, W.O. and Shimp, T.A. (1982), “ The use of ext rin-

sic cues to facili tate product adoption” , Journal of

Marketing Research, Vol. 19, May, pp. 229-39.

Birley, S. (1984), “ Finding the new firm” , in Pearce, J.A.

and Robinson R.B. (Eds), Proceedings of the 44th Annual Meeting of the Academy of Management,

Boston, MA.

Bryman, A. (1988), Quantity and Quality in Social

Research , Unwin Hyman, London.

Crosby, L.A., Evans, K.R. and Cowles, D. (1990), “ Relation-

ship quality in services selling: an interpersonal

influence perspective” , Journal of M arketing,

Vol. 54, pp. 68-81.

Deshpande, R. (1983), “ Paradigms lost: on theory and

method in research in marketing” , Journal of

Marketing, Vol. 47, Fall, pp. 101-10.

Eisenhardt, K.M. (1989), “ Building t heories from case

study research” , Academy of Management Review,

Vol.14 No. 4, pp. 532-50.

Fetterman, D.M. and Pitman, M.A. (Eds) (1986), Education-

al Evaluat ion: Ethnography in Theory, Practice, and

Politics, Sage Publications, Beverly Hil ls/London.

Good, D. (1988), “ Individuals, interpersonal relations and

trust” , in Ganbetta, D. (Ed.), Trust: Making and

Breaking Cooperative Relations, Basil Blackwell,

New York, NY.

Henwood, K.L. and Pidgeon, N.F. (1993), “ Qualitative

research and psychological theorising” , in Hammer-

sley, M. (Ed.), Social Research: Philosophy, Politics &

Practice, Sage Publications, London, pp. 14-33.

Kaplan H. (1964), The Conduct of Enquiry, Chandler, San

Francisco, CA.

McAllister, D.J. (1995), “Affect and cognition-based trust as

foundations for interpersonal cooperation in

organisations” , Academy of Management Journal,

Vol. 38 No. 1, pp. 24-59.

Mayer, R.C., Davis, J. and Schoorman, F.D. (1995), “ Anintegrative model of organisational trust” , Academy

of Management Review, Vol. 20 No. 3, pp. 709-34.

Mint zberg, H. (1979), “An emerging strategy of ‘di rect’

research” , Administrative Science Quarterly, Vol. 24,

pp. 580-9.

Moorman, C., Deshpande, R. and Zaltman, G. (1993),

“ Factors affecting trust in market research relation-

ships” , Journal of Marketing, Vol. 57, January,

pp. 81-101.

Neu, D. (1991), “ New stock issues and the institut ional

production of trust” , Accounting Organisations and

Society, Vol. 16 No. 2, pp. 185-200.

Neuman, W.L. (1997), Social Research Methods, Qualit a-

tive and Quantitative Approaches , Allyn & Bacon,

Needham Height s, MA.

Pettigrew, A. (1988), “ Longitudinal field research on

change” , paper presented at the National Science

Foundat ion Conference on Longit udinal Research

Methods in Organisations, Austin, TX.

Reichardt, C.S. and Cook, T.D. (1979), “ Beyond qualitative

versus quant itative methods” in Cook, T.D. and

Reichardt, C.S. (Eds), Qualitative and Quantitative

Methods in Evaluation Research, Sage, Beverly Hills,

CA, pp.7-32.

Rott er, J.B. (1971), “ Generalised expectancies for interper-sonal trust” , American Psychologist, Vol. 35, pp. 1-7.

Schwab, (1980), “ Construct validity in organisational

behavior” in Staw, M. and Cummings, L.L. (Eds),

Research in Organisational Behavior, JAI Press,

Greenwich, CT.

Sexton, D. (1987), “ Entrepreneurship research needs and

issues” in Sexton, D. and Smilor, R. (Eds), Entrepre-

neurship 2000, Upstart Publishing Company,

Chicago, IL.

Smilor, R. (1987), “ Preface” in Sexton, D. and Smilor, R.

(Eds), Entrepreneurship 2000, Upstart Publishing

Company, Chicago, IL.

Van de Ven, A.H. (1991), “ Longitudinal methods for

studying the process of entrepreneurship” in Sexton,

D.C. and Kasarda, J.D. (Eds), The State of the Art of

Entrepreneurship, P.W.S. Kent, Boston, MA.

Van Maanen, J. (1979), “ Reclaiming qualitative methods

for organisational research: a preface” , Administra-

ti ve Science Quarterly, Vol. 24, December, pp. 520-6.

Wiener, J.L. and Mow en, J.C. (1986), “ Source credibil ity: on

the independent effect of t rust and expertise” ,

Advances in Consumer Research, Vol. 13,

pp. 306-10.

Yin, R.K. (1984), Case Study Research: Design and

Methods , Sage, Beverly Hill s, CA.

110




Volume 2 · Number 2 · 1999 · 103–110

integrating deductive

Documents