Consumer Disposition toward Satisfaction (CDS): Scale Development and Validation

Author(s): Debra Grace

Source: Journal of Marketing Theory and Practice , Spring, 2005, Vol. 13, No. 2 (Spring,
2005), pp. 20-31
Published by: Taylor & Francis, Ltd.

Stable URL: [Link]

REFERENCES
Linked references are available on JSTOR for this article:
[Link]
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@[Link].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
[Link]

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marketing Theory and Practice

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
 CONSUMER DISPOSITION TOWARD
SATISFACTION (CDS): SCALE
DEVELOPMENT AND VALIDATION

Debra Grace
Griffith University

This paper conceptualizes the notion of "consumer disposition toward satisfaction" (CDS), defined as the consumer'
tendency to be sufficiently satisfied with the purchase and consumption of the goods and services they utilize. Furthermo
development and validation of the Consumer Disposition toward Satisfaction (CDS) Scale is detailed. The scale develo
process was carried out over three stages (item generation, scale purification, scale validation), which comprised of 8
data collections, involving a total of 1 5 1 4 individual respondents. As such, the data provide evidence for face, content, cr
related, discriminant and convergent validity, dimensionality, reliability and generalizability of the CDS scale. The C
is a unidimensional, parsimonious scale that has the potential for significant usage in the development and testing of t
well as in practical applications.

INTRODUCTION factors that influence satisfaction responses (e.g, consumer

characteristics), and these need to be considered when
Due to increased market-place competition interpreting
and buyersatisfaction data. For example, when asking a
sophistication within the past 30 years, onegroup
of the key (ranging in all ages) to rate their satisfaction
of people
emerging themes within the realms of marketing with
has been
a rockthat
concert it is more likely that younger people would
of customer satisfaction. During this time, therate
concept of higher than older people. Therefore, the
the concert
customer satisfaction has been the focus of considerable
results become more a product of age rather than of
academic deliberation and debate and, as such, theory satisfaction. If the effect of age was controlled for then a truer
development has been endemic in this area (Spreng, indication of actual satisfaction would emerge. However,
Mackenzie and Olshavsky, 1995). Furthermore, marketingwhile demographic information (such as age) is easily
practitioners, realising the importance of satisfying customers
measured and obtained, information regarding an individual's
over time, have also invested heavily in understanding and personality is often not. In fact, as personality characteristics
measuring customer satisfaction with the long-term goal ofare many and varied, measurement is often incomplete,
securing future profits and sustaining business viability laborious and, as a consequence, error-prone.
(Danaher and Haddrell, 1996). In fact, by 1994, the use of
customer satisfaction surveys had become so widely spread Acknowledging the inefficiencies of studying personality traits
that a review of the literature revealed 40 different satisfaction
in the consumer domain, Baumgartner (2002) proposes that
scales which had been used in various settings, ranging from rather than attempting to understand the consumer in terms of
retail to airlines (Danaher and Haddrell, 1996). their personality traits, we should get to know consumers as
dispositional entities in domain-specific contexts. For
However, the measurement of customer satisfaction has example, scepticism toward advertising is a dispositional
proven to be problematic (Petersen and Wilson, 1992) with
construct that relates specifically to one's tendency (to be
various reasons (many involving methodological issues) sceptical) in the domain of advertising. As such, it is not
having been cited for satisfaction rating anomalies. One well-
implied that the individual would be sceptical in all situations,
documented problem relates to the effect of consumer which would certainly be the case if the construct related to
characteristics on satisfaction ratings (Mittal and Kamakura,
personality (as opposed to a disposition), as personality
2001). This does not mean that the scale measurements for characteristics are more enduring and tend to be much more
satisfaction are necessarily flawed, but that there are other
consistent across situations. Therefore, by adopting a

20 Journal of Marketing THEORY AND PRACTICE

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
dispositional approach within the specific domain of customer wide zone of tolerance, in terms of their product expectations,
satisfaction, we can enhance our understanding of and account whereas those who are quite difficult to deal with have a very
for individual consumer differences in relation to satisfaction narrow zone of tolerance. This would explain differences in
judgements and evaluations of product offerings. It is on this expressed satisfaction ratings of consumers who have
basis that a new construct identified as "consumer disposition essentially had very similar product experiences. This notion
towards satisfaction" is proposed. As such, this paper was alluded to by Mittal and Kamakura (200 1 ) with regards to
discusses the conceptualization of consumer disposition satisfaction and repurchase intentions. They posited that
towards satisfaction (CDS) and the subsequent "consumers may have different thresholds or tolerance levels
towards repurchase" (p. 132) and that consumers with the
operationalization of this construct through the development
and validation of the Consumer Disposition to Satisfactionsame satisfaction rating may have different levels of
(CDS) Scale. repurchase behavior because of these differences. On this
basis, it could be suggested that some individuals are simply
CONCEPTUAL DEVELOPMENT predisposed to product satisfaction, whereas others are not.

Customer satisfaction has been conceptualized in many CONSUMER

ways DISPOSITION TOWARD
over the past 20 years. In fact, Oliver (1980), in building on SATISFACTION
Helsons (1948) adaptation level theory and Fishbein's (1967)
Zeithaml
attitude model, was instrumental in providing support foret
anal. (1993)'s notion of the zone of tole
integrated model of customer [Link]
Over time,an individual characteristic inherent
definitions of the construct have been many and varied,
consumers or, in other words, consumer tolerance in re
although consensus has, largely, been achieved on some
to product purchases. Given that consumer tolerance is
common elements. For example, it is generally toagreed that
be a function of what is expected (individual expecta
customer satisfaction is an emotive and/or cognitive
and response,
what is considered adequate (individual perception
pertaining to a particular focus (expectations, product
relation to product performance, then it would be expecte
performance), which occurs after purchase, choiceeachorindividual
product would have different tolerance levels,
experience has occurred (Giese and Cote, 2000).someMoreover,
individuals being more tolerant of product offering
customer satisfaction is generally understood others.
within the Therefore, it could be logically argued th
expectancy/disconfirmation paradigm (Zeithaml et al., 1993)
individual exhibiting greater tolerance in relation to p
whereby satisfaction/dissatisfaction is said to be the result
offerings of
would be more easily pleased (or satisfied) i
the comparison between pre-purchase expectations and than those who are less tolerant. This
marketplace
perceived product performance. However, because
othersvariability
have in product performance may be re
conceptualized customer satisfaction in terms of equity theory
acceptable to some individuals, yet rejected by others a
(Oliver and Swan, 1989), product performance a(Patterson,
result, this would directly impact on the indiv
1993; Churchill and Suprenant, 1982), desires congruency
consumer's level of satisfaction. This would explain
(Spreng et al., 1996; Spreng and Olshavsky, 1993),
someconsumer
consumers are easier to please than others.
value (Woodruff and Schumann, 1993) and consumption
emotion (Westbrook and Oliver, 1991). Given many different
However, while customers have long been categoris
conceptualizations within the literature, however, it"easily
labelled as is pleased" and/or "difficult", there has
generally agreed that customer satisfaction little
involves the attention given to this consumer tr
research
comparison of standards whether they be in the form of let alone any attempt to measure it (Moo
mannerism,
expectations, desires, wants, ideal or equitable performances.
and Olver, 1997). This is surprising given that pre
research has linked individual traits and/or disposition
In an attempt to address the diverse issues range surrounding
of consumption-related behaviors such as purcha
expectations and standards with regards to customersocietal influence, fear appeals, changes in at
risk-taking,
satisfaction, Zeithaml et al. (1993) proposed a comprehensive
(Kassarjian, 1971) and brand preference (Albanese,
Therefore,
model of expectations and their potential antecedents. if an efficient measurement in relatio
It was
within this framework they first proposed the consumer's
notion of the
overall tendency toward product satisfaction
zone of tolerance, which they describe as "the extent to which we could enhance our general understandi
be produced,
customers recognize and are willing to accept heterogeneity"
individual differences within consumer research and,
(Zeithaml et al., 1993, p. 6). It is on this basis they specifically,
argued that in the context of consumer evaluation
an individual's zone of tolerance is the difference between intentions. Furthermore, practical application of
what they desire and what they consider adequate, in terms ofmeasurement may well lead to more accurate interpretat
performance, and this zone can expand and contract acrosssatisfaction data through the standardisation of indiv
situations and individuals. This may explain why "somescores.
customers are consistently easy to please, [and] others are
interminably difficult" (Mooradian and Olver, 1997, p.389).The study of dispositions dates back to the wri
It may be that those customers who are easily pleased have aMunsterberg in 1913 and these are defined as "a

Spring 2005 21

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
customary manner of an emotional response" (Houghton of the goods and services they utilize. Given that CDS relates
Mifflin Company Dictionary 2005) or "an affective tendency" to an individual's customary manner of an emotional response
(Staw and Cohen-Charash, 2005). Even Roethlisberger in the context of consumption/purchase satisfaction, CDS is
( 1 94 1 ), as a result of the classic Hawthorne Studies, alluded to appropriately described as a consumer characteristic. As such,
the concept of dispositions in relation to job satisfaction. He CDS is not another expression of customer satisfaction, but
noted that those individuals who exhibited chronic job rather it is an integrative construct, which represents an
dissatisfaction may be "people whose dissatisfactions were individual consumer characteristic, thus exerting influence on
more deeply rooted in factors relating to their personal the consumer's cognitive and emotive reactions (customer
histories. For such people the simple remedy of changing the satisfaction) and behavioral intentions toward product
object of the complaint was not enough" (p. 1 8). Furthermore, offerings. Similar to constructs such as susceptibility to
Staw and Cohen-Charash (2005) argue that "a person's interpersonal influence (Bearden et al., 1989), need for
affective tendencies are not merely a means for researchers to emotion (Raman, Chattopadhyay and Hoyer, 1995) and
classify an individual's personality. They can also provide a scepticism toward advertising (Obermiller and Spangenberg,
heuristic by which others categorise and interpret an actor's 1998), CDS describes a general tendency that varies across
behavior, setting in motion a cycle of actions and reactions that individuals and is related to other individual traits and
may persist over time between actors and observers in a social characteristics.
environment" (p. 63). Moreover, in the context of consumer
behavior, Baumgartner (2002) advocates that one way in RESEARCH OBJECTIVE
which to overcome the inefficiencies in trait research is to
"develop an integrated framework for getting to know The goal of this research is to develop a theoreti
consumers as dispositional entities" (p. 287). For example, measurement scale for CDS, and, as such, consideration
while Baumgartner (2002) acknowledges that predictive be given to scale type and validity issues. On reviewin
accuracy can be efficient in the context of domain-specific scale development literature, Flynn and Pearcy (2001) n
individual difference variables, there has been a proliferation a common oversight in relation to scale development w
of trait measures, which, in many cases, only serve to confuse lack of recognition regarding the differences bet
rather than to clarify. This being the case, the need to "embed theoretical and applied scales and that the same dem
particular personality variables into more comprehensive should not be made of applied and theoretical measures
integrative frameworks" has arisen (Baumgartner, 2002, p. explain, "the ideal characteristics of a scale are not the
287). It is on this basis that a dispositional approach in if a researcher is measuring a construct or phenomeno
consumer research is advocated and, thus, adopted for this diagnostic or managerial reasons vs. measuring the
study. construct for theoretical explanation" (415). For exam
Flynn and Pearcy (200 1 ) advocate that if the research aim
While not nearly as common as trait-specific research in test theory by examining constructs in relation to ot
consumer behavior, over the past 20 years, dispositional phenomena then "no longer is an inventory of it
research has received considerable attention within the necessary.... we should be aiming for a concise list of
management literature, particularly in relation to capturing job the heart of the construct" (Flynn and Pearcy,
satisfaction (for a good summary of dispositional researchp. see418), rather than, in the case of applied scales, try
Davis-Blake and Pfeffer, 1989). in fact, Weiss and Adler achieve complete coverage of the construct in order to
(1984) argue that a truly dispositional approach supports accurate
the predictions. Furthermore, DeVellis (2003) sug
that, in general, shorter scales are more efficient as les
notion that individuals possess a set of traits that significantly
effect their attitudes and behavior (in all organizational
burden is placed on respondents. In terms of val
settings). Furthermore, Staw and Ross (1985) argued that a
construct validity is of major importance where theor
scales are concerned, therefore, tests for conver
predisposition toward happiness significantly effects job
discriminant
satisfaction while Pulakos and Schmitt ( 1 983) found "that high and nomological validity and inte
school seniors' expectations about the extent to which work
consistency should be entrenched in the scale develop
would satisfy existence, relatedness and growth needs wereprocess (Spector, 1992). Finally, "the theoretical scale
significantly positively correlated with job satisfactionto20 have a stable and known factor structure so theorist
months after graduation" (in Davis-Blake and Pfeffer, 1 989, p. on its consistent performance in complex models"
rely
391). On this basis, given that customer satisfaction and
is a Pearcy, 2001, p. 420). In terms of factor stru
function of product expectations just as job satisfaction because
is a a parsimonious measure of the CDS constr
function of job expectations, a dispositional approachsought,
to rather than the unique properties of its v
customer satisfaction may well be warranted. manifestations, a unidimensional scale (refer to Bagoz
Heatherton's (1994) total aggregation model of person
On the basis of the previous arguments, the concept of
construct measurement) is proposed. Unidimensionality
consumer disposition toward satisfaction (CDS) is proposeddictated by the nature of CDS given that it is an over
and defined, accordingly, as the consumer 's general tendency
general construct that encapsulates a consumer's tenden
to be sufficiently satisfied with the purchase and consumption
satisfaction over all purchasing and consumption situat

22 Journal of Marketing THEORY AND PRACTICE

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
 TABLE 1
As such, the research objective of this study is to development
VALIDITY AND RELIABILITY CONFIRMED
a concise unidimensional CDS scale that is both valid and
BY DATA COLLECTION
reliable and, furthermore, generalizable.

SCALE DEVELOPMENT State of Scale Source of Data Provides Evidence

Development for:
The methodology adopted for the development and testing of
Stage 1 Item Qualitative Data- Face Validity
the CDS scale was largely based upon the guidelines of
Generation Students (N = 75) Content Validity
DeVellis (2003), Churchill (1979) and Campbell and Fiske Expert Judges
( 1 959) with careful scrutiny being given to the methodological
strategies of widely-published scale development Stage 2 Survey One-Students Reliability (Internal
researchers,
Purification (N = 294) Consistency)
such as Bearden, Netemeyer, Richins, Goldsmith, Flynn, Dimensionality
Eastman, Sweeney and Soutar. Careful consideration in the
design of the scale development process was also given to the Survey Two-Consumers Reliability (Intern
four criticisms of scale development made by Flynn and (N = 312) Consistency)
Dimensionality
Pearcy (200 1 ) which involved inadequate development sample
sizes, insufficient replications, inconsistencyStage
in usage of
3 Validation Survey One (A/B)-Uni Dimens
exploratory factor analysis and issues concerning applied vs Students
theoretical scales. Therefore, the scale development process (N = 8 1 ) Reliability(Internal) Test/Retest
was carried out over three stages (refer Table 1), which Reliaability
comprised of 8 separate data collections, involving a total of (Temporal)
1514 respondents. The respondent pool included student Social Desirability
Bias
samples and general consumer samples. Student samples were
deemed to be appropriate and are acceptable in the scale Survey Two - Students Dimensionality
development process given that it is not a representative mean (N = 233) Reliability (Internal
value that is sought in the process but, more importantly, Consistency)
Discriminant
evidence of internal consistency of the scale (see DeVellis,
Validity
2003). For example, "a sample of this sort [student sample] Convergent Validity
might still lead to correct conclusions about which items are
most strongly interrelated" (DeVellis, 2003, p. 89). Survey Three- Dimensionality
Consumers Reliability (Internal
(N = 283) Consistency)
Data collected provide evidence for face validity, content Discriminant
validity, criterion-related validity, discriminant validity, Validity
convergent validity, dimensionality, and the reliability (both Convergent Validity
internal and temporal) of the CDS scale. Furthermore, a test
Survey Four- Dimensionality
for social desirability bias is undertaken and evidence of the
Consumers Reliability (Internal
generalizability of the CDS scale across consumer samples is (N = 236) Consistency)
provided. The ensuing discussion outlines the item generation, Discriminant

scale purification and scale validation stages of the CDS scale Validity
Convergent Validity
development process.

Stage One (Item Generation)

A qualitative approach, which included both literature search

removed from the conceptual themes identified in the literature
(and the
and open-ended questionnaires, was used to generate previously
initialdiscussed in theoretical development).
item pool. While the literature search, relating Therefore,
specifically to identified in the literature and generated
the themes
through
the theories of customer satisfaction, assisted in student responses
providing a (eg., expectations, desires and
selection of words and phrases to be used in thewants,
initial items,
product performance, feelings, value for money and
it was important to determine how potential respondents
satisfaction) provided the guide for item generation. This
viewed the concept of customer satisfaction. This being
resulted the encapsulating expectations (eg., On the
in 7 items
whole, to
case, a group of 75 university students were asked I have high expectations of the goods and services 1
write
down in their own words what was meant buy), by 7theitemstermrelating to desires and wants (e.g, I usually get
"customer satisfaction". The responses were content
exactly what
analysed
I want when I make my purchases), 7 items in
relation tothemes
and this resulted in the emergence of four predominant performance (e.g, When I buy goods, that don't
being, (1) having needs and wants being met (2) having
perform well, I get angry - reversed), 7 items relating to
feelings with
expectations of the product being met (3) being happy (e.g, Quite
oroften I am not happy with what I buy -
having positive feelings about the purchasereversed),
(4) receiving
7 items relating to value (e.g, I think I get value for
value for money. These themes were, in actual fact,
money not
with most of my purchases) and 7 items relating to

Spring 2005 23

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
satisfaction (e.g, On the whole, I would say I am easily Purification Sample 2:
satisfied when it comes to purchasing goods and services). In
addition, it was decided to use a 7-point Likert scale ranging The remaining 8 items were then included in a survey and
from "strongly disagree" to "strongly agree" due to its ability distributed to a broader sample of consumers (as opposed to
to measure opinions, beliefs and attitudes (DeVellis, 2003) and university students). As such, the data collection process
its common usage within the marketing domain. Given the involved a non-probability sampling process whereby students
decision to use a Likert response format, it was necessary, were asked to snowball the survey to friends and family. The
then, for each item to be presented as a declarative statement resulting sample comprised of 3 12 respondents ranging in age
(DeVellis, 2003). from 17 years to 63 years (mean age of 29 years) of which
68% were female. Gross annual income ranged from $5,000
As such, the initial pool of 42 items was generated and ready to $82,000 per annum with a mean of $56,000 per annum.
for fiirther scrutiny by experts (DeVellis, 2003). The initial
item pool was given to four independent judges (academics Firstly, the total sample was split randomly into sample 1
within the marketing domain) along with the constructs (n=149) and sample 2 (n=163). Exploratory factor analysis
definition (similar to Tian et al., 2001; Netemeyer and Teel, (using varimax rotation) was conducted on sample 1
1 989). At this point, the judges were asked to rate each of the producing a one-factor solution explaining 56% of the
items as either "not at all representative", "somewhat variance. However, there was one item (V7) that performed
representative" or "clearly representative" of the definition poorly with a factor loading of .39 and on inspection of
provided. Only those items that were judged as clearly reliability analysis its item-to-total correlation was < .50 and
representative of the definition, by at least three of the four this item also reduced alpha. As such, this item was deleted
judges, were retained. This resulted in 17 items being deleted. from further analysis. The remaining seven items were then
The remaining 25 items were considered to have sufficient factor analysed again which produced a one-factor structure
face and content validity and, as such, were ready to be explaining 67% of the variance (factor loadings ranged from
administered to a development sample for fiirther scale .76 to .91) and Cronbach's alpha was computed at .91.
purification.
Sample 2 was then used for the purposes of confirmatory
Stage Two (Scale Purification) factor analysis on the remaining 7 items. Amos 5.0 was sued
employing maximum likelihood estimation and the results
Purification Sample 1 : indicted a poor fit for the model. Examination of the
modification indices indicated that two error terms were highly
Data were collected on campus from second and third year correlated and an inspection of the corresponding standardized
university students in a large Australian university. The sample residual covariance (4.75) revealed that it was well in excess
totalled 294 students of which 42% were male and 58% were of the suggested cut-off of 2.58 (Byrne, 2001). On this basis,
female. Ages ranged from 17 years to 52 years with a mean and similar to the procedure of Li, Edwards and Lee (2002),
age of 23 years. Analysis included a combination of reliability these two items were dropped from further analysis. As such,
analysis with particular attention being given to corrected the confirmatory factor analysis was re-run on the remaining
item-to-total correlations and exploratory factor analysis. In five items which resulted in a good fit for the model with Chi-
accordance with the recommendations of Flynn and Pearcy square of 6.90 on 5 df, p = .228, goodness of fit index (GFI) of
(2001) and along similar lines to that used by Pecheux and .98, adjusted goodness of fit index (AGFI) of .95, root mean
Derbaix (1999), these two analytical procedures were used square residual (RMR) of .03 and root mean square error of
simultaneously. Initially, principal components factor analysis approximation (RMSEA) of .05 (refer results in Table 2). The
using oblique rotation (due to the expectation that any possible results provide evidence of unidimensionality and convergent
factors would be correlated) was conducted which resulted in validity as the GFI is above .95 and the AGFI is above .90,
a five-factor solution. There was one predominant factor while the RMR and the RMSEA is below .05 and .08
comprised of 1 3 items (eigenvalue of 5.97, explaining 33% of respectively (Chandon et al., 1997).
the variance) and four seemingly meaningless factors that
included the remaining 12 items which exhibited loadings and Stage Three (Validation)
cross-loadings of less than .5. These 12 items also exhibited
poor item-to-total correlations of less than .50 (cf. Validation Sample 1 :
Zaichkowsky, 1985) and, thus, were deleted from further
analysis. This procedure was then repeated another two times, The purpose of this data collection phase was to examine the
which resulted in the deletion of 5 items and 2 items CDS scale in terms of its temporal stability using a test-retest
respectively. The final result was a uni-dimensional factor
reliability procedure and to assess social desirability bias. As
structure explaining 56% of the variance, which included
such,8the survey included the five CDS items, 10 items from
items exhibiting factor loadings ranging from .65 to .81 and
the Marlowe-Crowne (MC) 2 social desirability scale (Stahan
and
the computed reliability estimate of Cronbach's alpha was Gerbasi, 1972) and, although not included for data
.84
(refer Table 2). analysis purposes, random items from the CD scale were

24 Journal of Marketing THEORY AND PRACTICE

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
 TABLE 2
PSYCHOMETRIC PROPERTIES OF CDS SCALE - RESULTS ACROSS ALL DATA COLLECTIONS

Data Purification Purification Validation 1 Validation 2 Validation 3 Validation 4

Collections 1 2 (2 Collections)

Respondents Unit Students Consumers Uni Students Uni Students Consumers Consumers
N= 294 312 81 233 283 236

SUMMED CDS A B
Mean 26.46 26.95 26.10 27.10 24.80 25.71
St Deviation 4.62 4.35 4.54 4.57 4.76 4.37

EFA (8 Items) (n-149) A B

Variance Exp. 56% 67% 67% 76%
Loadings .65-81 .76-.91 .78-.91 .81-92

CFA (N=163)
Chi-Square 6.90 10.24 11.42 6.43
df 5 5 5 5
pValue .228 .106 .044 .266
GFI .98 .98 .98 .98
AGFI .95 .95 .95 .96
RMR .03 .03 .07 .02
RMSEA .06 .06 .06 .04

RELIABILITY
Cron Alpha .84 .88 .85 .85 .88 .78 .87
Test/Retest

.75

BIAS
Social .024 .10
Desirability
TABLE 3

Consumer Disposition toward Satisfaction (CDS) Items Loading

Usually I am pleased with what I buy. .87

More often than not, I am a satisfied consumer. .83

I generally find the goods and services I buy don't live up to my expectations. ♦♦ .82

Quite often I am dissatisfied with my purchases of good and services. ** .82

Overall, 1 am usually satisfied with the purchases I make. .77

*♦ Indicates reverse scoring

scattered throughout the survey in order to break-up

in the the other
latent construct and exhibiting fac
scales and provide reasonable length from
to the.78
survey.
to .91Data
and were
for the second sampl
collected on campus on two separate ofoccasions (four
the variance weeks
with factor loadings rang
apart) from the same student respondents from
In terms oftwo different
reliability, average inter-item
Australian universities (1 private and
university andalpha
coefficient 1 public
was .85 for sample A
university). One hundred and fifty-seven surveys
correlations were .70were
with coefficient alp
[Link],
collected in the initial data collection, The correlation
of this between
initial the summed C
group, only 81 students completed the
data second survey,
collections was .75 which suppor
representing a response rate of 52%. The sample
reliability of the (n=81)
scale and is comparable
comprised of 27 males and 54 females scale
ranging in age from
test/retest 1 8
reliabilities (eg., Flynn et
to 41 years, with a mean age of 26 years.
Goldsmith, 1999; Tian et al, 2001). On thi
consistency and temporal stability of
confirmed.
Exploratory factor analysis (using varimax rotation) of both
samples supported the unidimensionality of the scale with the
results for the first sample (A) explaining 67% of the variance

Spring 2005 25

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
In order to determine if social desirability was a potential Validation Sample 3:
confound, the summed CDS scale and the five individual CDS
items were correlated with the social desirability scale. The The purpose of this data collection was to further examine
results indicated there were no significant correlations with r dimensionality, reliability and convergent and discriminant
= .024 being reported for sample A and r =. 1 0 being reported validity over a more diverse consumer sample. As such, the
for sample B. As such, social desirability was not considered survey included items measuring Life Satisfaction (SWLS)
to be problematic for the CDS measures. Refer Table 2 for (Diener, 1984), Self-esteem (Rosenberg, 1986), Cognitive
summary of results. Dissonance (Sweeney et al., 2000) as these were all constructs
that would be expected to be relate, in some manner, to CDS.
Validation Sample 2: Data were collected via on-line email survey from 283
consumers aged between 1 6 and 63 years (mean age 32 years),
The purpose of the second validation sample was to further 35% of which were male and 65% were female and annual
assess the dimensionality, reliability and construct validity of incomes ranged from $15,000 to $200,000 per annum with a
the CDS scale. In order to achieve this, the survey for this data mean of $3 1,000 p.a.
collection included a 5-item Satisfaction with Life Scale
(SWLS) (Diener et al., 1985), 5 items from Rosenberg's Confirmatory factor analysis revealed a good fit for the
(1986) Self-Esteem Scale, the 5-item CDS scale and measurement model with a Chi-square of 1 1.42 on 5df (p =
demographic items. Data were collected from a convenience.044), GFI = .98, AGFI = .95, RMR = .07 and RMSEA = .06.
sample of 233 university students aged 18 to 47 years with a Furthermore, the internal consistency of the CDS scale was
mean age of 25 years. confirmed by the reliability estimate of Cronbach's alpha .78
(refer Table 2). In addition, in this sample, discriminant
Confirmatory factor analysis, which was conducted via AMOSvalidity was evaluated by a similar procedure to O'Cass
5.0 using maximum likelihood estimation resulted in a Chi-(2002) who argued that evidence of discriminant validity
square of 10.24 of 5 df and the p-value was .106. The GFI exists if the correlation between two composite constructs is
(.98) and the AGFI (.95) were within acceptable levels, as was not higher than their respective reliability estimates.
the RMR (.03) and RMSEA (.06), thus indicating a good fit forCorrelations ranged from .40 to .56 and the reliability
the measurement model and confirming the unidimensionality estimates ranged from .75 to .86, thus providing evidence of
and convergent validity of the CDS scale. Furthermore, the discriminant validity.
reliability for the CDS scale was calculated via Cronbach's
alpha and confirmed at .88 (refer Table 2). The CDS scale was further examined to determine if the
summated CDS scale performed in the expected manner
Discriminant validity was evaluated via a similar process towithin a theoretically based model. Firstly, as previously
Eastman et al (1999) whereby items relating to other discussed, it would be expected that CDS would be positively
constructs (eg. overall life satisfaction and self-esteem),related to overall life satisfaction and, this being the case,
considered distinct from, but closely related to, the CDS scaleshould therefore be positively related to self-esteem (given the
were examined via exploratory factor analysis. The SWLS wasestablished relationship between life satisfaction and self-
chosen because it provides a measure of overall lifeesteem eg. Unsal, 2003). On this basis, it was hypothesized
satisfaction (Diener, 1984) and it would be expected thatthat there would be a significant positive relationship between
satisfaction with product purchase would be related to overallCDS and life satisfaction and a significant positive
life satisfaction, but within the specific domain ofrelationship between CDS and self-esteem. Bivariate
consumption. On this basis, it would be expected that, correlations supported these hypothesized relationships with
although closely related to life satisfaction, it would be a Pearson's correlation between CDS and life satisfaction being
distinct construct within its own right. Similarly, because of.56 and between CDS and self-esteem the correlation was .40
the close relation previously found between self-esteem and (p<.001 in both cases).
life satisfaction (eg., Unsal, 2003; Schmitt et al, 1979), it
would be expected that CDS would be distinct from, butFurthermore, in behavioral terms, and as previously discussed,
closely related to self-esteem. The factor analysis results it would be expected that someone who is generally quite
indicated a clear three-factor structure, which explained 72%satisfied with the purchases they make would, overall, be less
of the variance and there were no cross loadings greater thanlikely to experience cognitive dissonance (CD). On this basis,
2.0. The first two factors (life satisfaction and CDS) hadit would be expected that a negative relationship would exist
eigenvalues greater than 2.0, while self-esteem made up the between CDS and CD. This relationship was supported by the
third factor with an eigenvalue of 1.21. On this basis, the results of correlation analysis which revealed a significant
discriminant validity of the CDS scale was confirmed as nonegative correlation of -.56 where p < .01. Given that the
items from different scales loaded on the same factor (Lusch,results of this data collection support the three relationships
1976). proposed, further evidence of construct validity of the CDS
scale was provided, the results of which can be found in Table
4.

26 Journal of Marketing THEORY AND PRACTICE

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
 TABLE 4
behavior (eg., McCarthy, 2000, Mitchell, 1999), Previous
RESULTS CONSTRUCT VALIDITY TESTS (CORRELATIONS)
research suggests that individuals who are willing to take risks
engage in activities or situations where the possible outcomes
Validation Cognitive are generally poor or negative (Mitchell and Boustani, 1993).
Sample 3 Life Satisfaction Self-Esteem Dissonance If this is the case then, in terms of consumer behavior, it is
more likely that the consumption decisions of risk-takers will
CDS .56** .40** -.56**
lead to negative outcomes over and above the decisions of
Validation Status individuals who are risk adverse. Therefore, it would be
Sample 4 Perceived expected that, generally, risk-takers
Control would be more likely to ^^
Taking on experience dissatisfaction with their consumption decisions
than would non-risk-takers. On this basis, it would be expected
.24** -.25** -.16* that a significant negative relationship would exist between
CDS risk-taking and CDS. Moreover, the results of correlation
analysis supported this proposition with Pearson's r = -.25 and
* Correlation significant at the 0.05 level p<.01.
** Correlation significant at the 0.01 level
Finally, the relationship between status consumption and CDS
was examined. Status consumption is defined as " the
The final data collection was undertaken to further examine
motivational process by which individuals strive to improve
the psychometric properties of the scale, establish convergent
and discriminant validity and provide further evidencetheirof social standing through the conspicuous consumption of
construct validity across a second consumer [Link]
The products that confer and symbolize status both for
survey included items measuring Status Consumption the individual and surrounding significant others" (Eastman et
(Eastman et al, 1999), Risk-taking (Raju, 1980), Locus al, of
1999, p. 42). Given that Eastman et al. (1999) found status
consumption to be positively related to materialism, and
Control (James, 1957) and which were expected to have some
relationship with CDS. Data were collected frommaterialism
236 has been consistently found to be negatively
consumers using a convenience sample who were aged related to life satisfaction (eg., Belie, 1984, 1985; Dawson and
Bamossy, 1991, Kau et al, 2000), it would be expected that a
between 16 and 68 years (mean of 42 years), with annual
negative relationship may well exist between status
incomes ranging between $ 14,000 p.a. and $89,000 p.a. (mean
of $39,000 p.a.) and 30% of the sample was male, while consumption
70% and CDS. This proposition was supported by the
were female. results of correlation analysis, which revealed that a significant
negative relationship (r = -. 1 6 p < .05) did exist between status
consumption and CDS. As such, the three relationships
Confirmatory factor analysis established the unidimensional
examined
nature and convergent validity of the scale with Chi-square of within this data collection were found to be as
6.43 on 5df (p = .266), GFI = .98, AGFI = .96, RMR = .02expected,
and thus providing further evidence of construct validity
RMSEA = .04. In addition, Cronbach's alpha of .87 was (refer results in Table 4.)
computed thus confirming the reliability of the scale (refer
Table 2). Discriminant validity was also assessed via test DISCUSSION AND IMPLICATIONS
recommended by O'Cass (2002) by comparing the composite
construct correlations with construct reliabilities and, asIn
the
accordance with the research objective of this study,
correlations (ranging from .16 to .25) were less thanscale the development and validation process undertaken for
reliabilities (ranging from .75 to .92), discriminant validityCDS
was scale has resulted in a parsimonious, unidimens
confirmed. scale which, over a series of separate studies, h
demonstrated its validity, reliability and generalizability
suggested that the key strength of the CDS scale lies i
In order to initiate a further investigation of construct validity,
simplicity as it can be used as an efficient mea
relationships were tested between CDS and perceived control,
measurement
risk-taking and status consumption. Firstly, given that self- when examining CDS within both theore
esteem and perceived control (or, in other words, those and
who practical domains and, as such, these are furt
have an internal locus of control) have previously found discussed.
to be
positively related (eg., Skinner, 1996), it would be expected
The theoretical implications of the development of the CDS
that CDS may also have a positive relationship with perceived
scale are numerous. Firstly, CDS should be further examined
control. In order to test for this relationship, Pearson's
correlation was computed and this was significant at .24in(p<relation to consumer-related personality and demographic
.01), thus indicating that the relationship found was as
variables in an attempt to provide a personology of the type of
expected. consumer who is generally happy with their consumption
decisions (or has a wide zone of tolerance). As a result, the
Secondly, in terms of personality characteristics, risk-taking identification of certain segments of consumers who are
has often been argued to affect consumer and marketplace predisposed to satisfaction or, more importantly,
dissatisfaction, may be possible. Secondly, CDS could be

Spring 2005 27

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
examined across different product types by adapting wording In contrast, when low CDS consumers expre
to specific product categories. For example, CDS could be ratings and high CDS consumers express h
compared across high and low involvement products or ratings, it could mean that these satisfaction
professional and non-professional services to determine the certain extent, reflect the consumer's disposi
extent to which product type influences customer satisfaction. to the former, while it should be assumed th
Similarly, we can also enhance our understanding of consumer inadequate, it may not necessarily be d
behavior within a global marketplace by examining CDS corrective actions should be taken pending
scores across respondents from different cultures. regards to the latter, it could be assumed tha
at minimum, adequate, but not necessarily
Although the CDS scale was developed, primarily, for such, there may still be considerable room f
theoretical purposes, it is suggested that, due to the efficiency
of the scale, it may well be useful in applied settings. In Finally, it is suggested that the wording o
addition, within the scale development process, the scale could be adapted to reflect the consume
exhibited good content and predictive validity and test-retest satisfaction in relation to a specific goods/
reliability, which is important for applied scales (Nunnally, For example, in examining the perceived p
1967). As such, it is suggested that the CDS scale could be banks, item wording could be changed
used as a tool to assist in the interpretation of customer "usually, I am pleased with banking services
satisfaction ratings through the comparison of CDS scores with I am dissatisfied with banks"). This propos
expressed satisfaction scores, thus providing richer the basis that a consumer's involvement o
information regarding product performance. with a particular product category may
disposition to satisfaction in that product/
An example of such usage of the CDS scale is provided in Therefore, someone who is generally h
Table 5. Firstly, optimal performance of a product (whether consumption experiences, yet has experien
goods or services) would be signified by high expressed negativity when dealing with, for example,
satisfaction ratings of those consumers who have low CDS may exhibit lower CDS scores for that prod
scores. Such scores would indicate that those individuals, who this sense, expressed satisfaction toward a
are generally hard to please, are highly satisfied with the could then be interpreted in relation the cons
product and, as such, it would suggest that current product to satisfaction with competitors serving the
strategy was highly effective and, therefore, should be
reinforced CONCLUSION

On the other hand, if low satisfaction ratings are received from In departing from the traditional personality trait approa
consumers who rate highly in terms of CSD, then dire understanding consumers and their subsequent behaviors
performance of the product could be assumed. In this sense, research adopts an integrated approach, recommended
consumers who are generally easily pleased, are dissatisfied Baumgartner (2002), which treats "consumers as disposi
with the product offering. This would indicate that current entities". In doing so, the concept of consumer disposi
product strategy was severely flawed and corrective actions toward satisfaction (CDS) is proposed and defined as
were essential in order ensure the longevity of the product. consumer 's general tendency to be sufficiently satisfied
the purchase and consumption of the goods and services
TABLE 5
utilize. Furthermore, a short unidimensional scale to me
INTERPRETING SATISFACTION RATINGS
ACCOUNTING FOR CDS CDS has been developed over a three-stage scale developm
process, which has provided substantial evidence of the
Dimension Expressed Satisfaction Expressed Satisfaction
scale's reliability, validity and generalizability. As such,
Low High CDS scale has the potential to make a considera
contribution in both theoretical and practical terms and, d
Consumer Disposition Inadequate Performance Optimal Perfor
to Satisfaction {Corrective actions (Reinforce current its parsimonious nature, may prove to be a valuable diagn
Low pending assessment) practice) tool for academicians and marketers in the future.

Consumer Disposition Dire Performance Adequate Performance

to Satisfaction {Corrective actions (Optimal performance
High essential) cannot be assumed)

28 Journal of Marketing THEORY AND PRACTICE

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
 REFERENCES

Davis-Blake,
Albanese, P. J. (1990), "Personality, consumer behavior and Alison and Jeffrey Pfeffer (1989), "Just a
marketing research" In E. C. Hirschman (Ed.),
mirage: The search for dispositional effects in
Research in Consumer Behavior (Vol. 4, pp. 1-49).organisational research," Academy of Management:
Greenwich, Conn: JAI Press, Inc. The Academy of Management Review, 14 (3), 385-
400.
Bagozzi, Richard P. and T.F. Heatherton (1994), "A general
approach to representing multifaceted personality
Dawson, S. and G. Bamossy (1991), "If "we are what we are,
constructs: Application to state self-esteem,"what are we when we don't have?: An exploratory
Structural Equation Modelling, 1, 35-67. study of materialism among expatriate Americans,"
Journal of Personality and Social Behavior, 6, 363-
Baumgartner, Hans (2002) "Toward a personology of 384.
the
consumer," Journal of Consumer Research 29
(September), 286-292. Houghton Mifflin Company Dictionary (online)
[Link]
Bearden, William O., Richard G. Netemeyer and Jesse E. Teel
Diener,to
(1989), "Measurement of consumer susceptibility Ed, Robert A. Emmons, Randy J. Larsen and S.
interpersonal influence," Journal of Consumer Griffin (1985), "The satisfaction with life scale,"
Research 15 (March), 473-481. Journal of Personality Assessment, 49 (1), 71-75.

Belk, Russell W. (1984), "Three scales to measure DeVellis,

constructs Robert F (2003), "Scale Development: Theory and
related to materialism: Reliability, validity, and Applications 2nd Edition. Applied Social Research
relationships to measure of happiness," Advances in Methods Series Volume 26" Thousand Oaks, CA,
Consumer Research 1 1, 291-297. Sage Publications.

Byrne, Barbara M. (2001), "Structural EquationEastman,

Modeling
Jacqueline K., Ronald E. Goldsmith and Leisa
with AMOS: Basic concepts, applications and
Flynn (1999), "Status consumption in consumer
programming," Mahwah, New Jersey, Lawrence
behavior: Scale development and validation,"
Erlbaum Associates Publishers. Journal of Marketing Theory and Practice, 7 (3),
41-52.
Campbell, Donald T. and Donald W. Fiske (1959),
"Convergent and discriminant validation by the Flynn, Leisa R., Ronald E. Goldsmith and Jacqueline K.
multitrait-multimethod matrix," Psychological Eastman (1996), "Opinion Leaders and Opinion
Bulletin, 56, 81-105. Seekers: Two New Measurement Scales," Journal
of the Academy of Marketing Science, 24 (2), 137-
Chandon, Jean-Louis, Pierre- Y ves Leo and Jean Philippe 147.
(1997), "Service encounter dimensions - a dyadic
perspective: Measuring the dimensions of service Flynn, Leisa R. and Ronald E. Goldsmith (1999), "A short,
encounters as perceived by customers and reliable measure of subjective knowledge," Journal
personnel," International Journal of Service Industry of Business Research, 46, 57-66.
Management, 8 (1), 65-76.
Flynn, Leisa R. and Dawn Pearcy, (2001), "Four subtle sins
Churchill, Gilbert A. Jr. and Carol Suprenant (1982), "An in scale development: Some suggestions for
investigation into the determinants of customer strengthening the current paradigm," International
satisfaction." Journal of Marketing Research, 19 Journal of Market Research, 43(4), 409-433.
(November), 491-504.
Gerbing, David W. and James C. Anderson (1988), "An
Churchill, Gilbert A. Jr. (1979), "A paradigm for developing updated paradigm for scale development
better measures of marketing constructs," Journal of incorporating unidimensional ity and its
Marketing Research XVI (February), 64-73. assessment," Journal of Marketing Research, 22,
11-19.
Danaher, Peter J. and Vanessa Haddrell (1996), "A
comparison of question scales used for measuring
customer satisfaction," International Journal of
Service Industry Management, 7 (4), 4-26.

Spring 2005 29

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
Giese, Joan L. and Joseph A. Cote (2000), "Defining O'Cass, Aron (2002), "Political advertising believability and
customer satisfaction," Academy of Marketing information source value during elections," Journal
Science Review [Online] 2000 (1), Available: of Advertising, 31 (1), 63-74.
[Link] 1 -
[Link] Oliver, Richard L. and John E Swan (1989), "Postpurchase
communications by consumers." Journal of
Hair, Joseph E., Rolph E. Anderson, Ronald L. Tatham and Retailing, 64(4), 516-533.
William C. Black (1998), "Multivariate Data
Analysis (Fifth Edition)" Upper Saddler River, Patterson, Paul G. (1993), "Expectations and product
New Jersey, Prentice Hall. performance as determinants of satisfaction for a
high-involvement purchase," Psychology &
Kassarjian, H. H. (1971), "Personality and consumer Marketing, 10 (5), 449-465.
behavior: A review," Journal of Marketing
Research 8 (November), 409-418. Pecheux, Claude and Christian Derbaix (1999), "Mood and
children: Proposition of a measurement scale,"
Kau, Ah Keng, Jung Kwon, Tan Soo Jiuan and Jochen Wirtz Journal of Economic Psychology, 20 (5), 571-590.
(2000), "The influence of materialistic inclination
on values, life satisfaction and aspirations: An Peterson, Robert A. and William R. Wilson (1992),
empirical analysis," Social Indicators Research, 49, "Measuring customer satisfaction: Fact or
317-333. artifact?", Journal of the Academy of Marketing
Science, 20(1), 61-71.
Li, Hairong, Steven M. Edwards and Joo-Hyun Lee (2002),
"Measuring the intrusiveness of advertisements: Pulakos, Elaine D. and Neal Schmitt (1983), "A longitudinal
Scale development and validation," Journal of study of a valence model approach for the
Advertising, XXXI (2), 37-47. prediction of job satisfaction of new employees,"
Journal of Applied Psychology, 68, 307-312.
Lusch, Robert F. (1976), "Sources of power: Their impact
on intrachannel conflict," Journal of Marketing Raman, Nirijan V., Prithviraj Chattopadhyay and Wayne D.
Research, XIII (November), 382-390. Hoyer (1995), "Do consumers seek emotional
situations: The Need for Emotions Scale," In Frand
McCarthy, Breda (2000), "The cult of risk taking and social Kardes and Mita Sujan (Eds), Advances in
learning: A study of Irish entrepreneurs," Consumer Research (Vol. 22, pp. 537-542). Provo,
Management Decision, 38 (8), 563-574. UT: Association for Consumer Research.

Mitchell, V-W. (1999), "Consumer perceived risk: Ray, John J. (1985), "Acquiescence and response skewness
conceptualisations and models," European Journal in scale construction: A paradox," Personality and
of Marketing, 33 (1/2), 163-195. Individual Differences, 6, 655-656.

Mitchell, V-W and P. Boustani (1993), "Market Roethlisberger, F.J. (1941), Management and Morale,
development using new products and new Cambridge, MA: Harvard University Press.
customers: A role for perceived risk," European
Journal of Marketing, 27 (2), 17-32. Rosenberg, Morris (1986), Conceiving the Self Malabar,
Florida: Robert E. Krieger Publishing Company.
Mittal, Vikas and Wagner A. Kamakura (2001),
"Satisfaction, repurchase, intent, and repurchase Schmitt, Neal, Kenneth J. White, Bryan W. Coyle and John
behavior: Investigating the moderating effect of Rauschenberger (1979), "Retirement and life
customer characteristics," Journal of Marketing satisfaction," Academy of Management Journal, 22
Research XXXVIII (February), 131-142. (2), 282-291.

Mooradian, Todd A. and James M. Olver (1997), "I can't Skinner, Ellen (1996), "A guide to constructs of control,"
get no satisfaction: The impact of personality and Journal of Personality and Social Psychology, 71
emotion on postpurchase processes," Psychology & (3), 549 -570.
Marketing, 14 (4), 379-392.
Spreng, Richard A., Scott B. MacKenzie and Richard W.
Obermiller, Carl and Eric Spangenberg (1998), Olshavsky (1996), "A Reexamination of the
"Development of a scale to measure scepticism determinants of consumer satisfaction." Journal of
toward advertising," Journal of Consumer Marketing, 60 (July), 15-32.
Psychology, 7 (2), 159-186.

30 Journal of Marketing THEORY AND PRACTICE

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]
Spreng, Richard A. and Richard W. Olshavsky (1993), "A Weis, H.M. and S. Adler (1984), "Personality and
desires congruency model of consumer organizational behavior," In B.M. Staw and L.L.
satisfaction," Academy of Marketing Science Cummings (Eds). Research in Organizational
Journal 21 (3), 169-177. Behavior. Greenwich, CT: JAI Press.
Staw, Barry M. and Yochi Cohen-Charash (2005), "The
dispositional approach to job satisfaction: more Westbrook, Robert A. and Richard L. Oliver (1991), "The
than a mirage, but not yet an oasis," Journal of dimensionality of consumption emotion patterns
Organizational Behavior, 26, 59-78. and consumer satisfaction," Journal of Consumer
Research, 18 (June), 84-91.
Staw, Barry M. and Jerry Ross (1985), "Stability in the
midst of change: A dispositional approach to job Widing, Robert, Jagdish N. Sheth, Sue Pulendran, Banwari
attitudes," Journal of Applied Psychology, 70, 469- Mittal and Bruce I. Newman (2003), Customer
480. Behavior: Consumer Behavior and Beyond,
Southbank Victoria: Thomson Learning Australia.
Strahan, R. and K.C. Gerbasi (1972), "Short, homogeneous
versions of the Marlowe-Crowne social desirability Woodruff, Robert B. and David W. Schumann (1993),
scale," Journal of Clinical Psychology, 28, 191- "Understanding value and satisfaction from the
193. customer's point of view," Survey of Business, 29
(1), 33-42.
Sweeney, Julian C, David Hausknecht and Geoffrey N.
Soutar (2000), "Cognitive dissonance after Zaichkowsky, Judith Lynne (1985), "Measuring the
purchase: A multidimensional scale," Psychology involvement construct," Journal of Consumer
& Marketing, 17 (5), 369-385. Research, 12 (3), 341-352.

Tian, Kelly T., William O. Bearden and Gary L. Hunter Zeithaml, Vaiane A., Leonard L. Berry and A. Parasuraman
(2001), "Consumers' need for uniqueness: scale (1993), "The nature and determinants of customer
development and validation," Journal of Consumer expectations of service," Journal of the Academy of
Research, 28 (June), 50-66. Marketing Science, 21 (Winter), 1-12.

Unsal, Yetim (2003), "The impact of

individualism/collectivism, self-esteem, and feeling
of mastery on life satisfaction among Turkish
university students and academicians," Social
Indicators Research, 61 (3), 297-318.

AUTHOR BIOGRAPHY

Debra Grace (Ph.D., Griffith University) is a senior lecturer in marketing at Griffith University, Gold
research and teaching interests lie within the services marketing, branding and consumer behavior a
a number of publications within journals that have their focus in these areas such as Journal of
European Journal of Marketing, Service Marketing Quarterly and others.

Spring 2005 31

This content downloaded from

[Link] on Mon, 24 Feb 2025 [Link] UTC
All use subject to [Link]