Journal of Retailing 79 (2003) 7795

Hedonic shopping motivations

Mark J. Arnold a,, Kristy E. Reynolds b
a
Department of Marketing, John Cook School of Business, Saint Louis University, 3674 Lindell Blvd., St. Louis, MO 63108, USA
b
Department of Marketing, E.J. Ourso College of Business Administration, Louisiana State University, Baton Rouge, LA, USA

Accepted 12 February 2003

Abstract

Given the increasing importance of entertainment as a retailing strategy, this study identifies a comprehensive inventory of consumers
hedonic shopping motivations. Based on exploratory qualitative and quantitative studies, a six-factor scale is developed that consists of
adventure, gratification, role, value, social, and idea shopping motivations. Using the six-factor hedonic shopping motivation profiles, a
cluster analysis of adult consumers reveals five shopper segments, called here the Minimalists, the Gatherers, the Providers, the
Enthusiasts, and the Traditionalists. The utility of the proposed scale is discussed both for future research and retail strategy.
2003 by New York University. Published by Elsevier Science. All rights reserved.

Keywords: Hedonic shopping; Scale development; Shopping motivations; Hedonic consumption; Retail strategy; Shopper segments

Over the years, retailers have been buffeted by a num- landscape the hedonic experiences that a customer can now
ber of macro-environmental forces that have changed the obtain are virtually endless: from rock-climbing walls in
landscape of the industry. These include the spread of mass shoe stores, to singles nights in grocery stores, to off-
discounters, the proliferation of suburban power centers road test tracks in Land Rover dealerships (e.g., Fournier,
and lifestyle retailing formats, and the recent arrival of the 1996). While retailers are focusing more on entertainment,
Internet as an alternative retail platform offering consumers academic research is lagging in investigating the hedonic
unparalleled convenience. For example, the July 1998 reasons people go shopping. For example, the last compre-
cover of Time magazine predicted the demise of the hensive effort at examining shopping motivations occurred
shopping mall: Kiss Your Mall Good-Bye: Online some time ago (Westbrook & Black, 1985), and the retail
Shopping is Cheaper, Quicker and Better. landscape has changed dramatically since then. Recent re-
In this environment it is no longer enough for a retailer tail research is beginning to focus on the hedonic aspects
to operate in a conventional manner by enticing customers of the in-store experience, such as the affective response of
with broad assortments, low pricing, and extended store excitement (Wakefield & Baker, 1998). However, no recent
hours. The entertainment aspect of retailing, or research has investigated, in a comprehensive manner, the
entertailing, is in- creasingly being recognized as a key multiple and varied hedonic reasons, or motivations, that
competitive tool. Many retailers are responding to the people go shopping. Therefore, given the current focus by
threat of Internet-based shop- ping by leveraging the retailers on the hedonic aspects of shopping and the general
brick-and-mortar advantages that virtual retailers cannot lack of academic activity in this area, there is clearly a need
match: higher levels of service, highly trained staff, and an for research on this issue.
entertaining and fun retail environment (Burke, 1997;
This study investigates the hedonic reasons people go
Cope, 1996; Wakefield & Baker, 1998). Re- tailers from
shopping. Based on qualitative and quantitative studies, a
supermarkets to video stores are sporting new and exciting
scale that measures hedonic shopping motivations is de-
ideas, such as animatronic farm animals, butter churning
veloped and validated. Simply put, a sound measurement
contests, and roaming face painters and childrens
instrument provides a foundation for future research inves-
performers (Buss, 1997). In fact, in this evolving retail
tigating the interrelationships between hedonic motivations,
in-store experiences, shopping outcomes (e.g., satisfaction),

Corresponding author. Tel.: +1-314-977-3612 (Office)/3868 (Depart- and specific shopping behaviors such as impulse buying.
ment); fax: +1-314-977-1481.
E-mail address: [email protected] (M.J. Arnold). Further, retailers would have a tool that could be employed
to
0022-4359/03/$ see front matter 2003 by New York University. Published by Elsevier Science. All rights reserved.
doi:10.1016/S0022-4359(03)00007-1
78 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

examine current and potential patrons, thereby providing shoppers have previously described themselves as a kid in
guidance for store design and marketing communications a candy store when engaged in holiday shopping, often
strategy. ex- pressing excitement, increased arousal, and a deep
Therefore, the major objectives of this research include: sense of enjoyment in shopping for others (Fischer &
1. Qualitatively investigate the hedonic reasons people go Arnold, 1990,
shopping; p. 334). Informants have expressed a sense of escapism
2. Develop and purify a scale measuring hedonic shopping while shopping, often describing the shopping trip as an
motivations; adventure: Shopping is . . . an adventure. When you cant
3. Validate the hedonic shopping motivations scale on a or dont find [what youre after] its o.k. because there are
separate sample of shoppers; lots of other places to look (Babin et al., 1994, p. 646).
4. Construct a taxonomy of shoppers based on their Shoppers have also described the enjoyment of bargaining
hedonic shopping motivations. and haggling (Sherry, 1990) and the mood-altering qualities
of the shop- ping experience (Thompson, Locander, &
The remainder of this paper is divided into four major Pollio, 1990). In summarizing these aspects of shopping,
sections. First, we discuss the theoretical background and Sherry (1990) con- cludes that the seeking of such
previous research that has been conducted in this area. Sec- experiences is often far more significant than the mere
ond, we present the results of a qualitative investigation acquisition of products (p. 27; see also Babin et al., 1994).
into hedonic shopping motivations, and discuss our
findings in relation to existing theory. Based on this we Shopping motivations
then develop an initial pool of scale items. Third, we
present the results of a multi-sample investigation that
Early studies developed taxonomies of retail shoppers,
serves to purify and validate the hedonic shopping
of- ten in an attempt to infer shopping motivations from
motivation scale. Finally, we provide a general discussion
distinct types of shoppers, such as the economic, or
of the findings, as well as limitations of the study and
apathetic shopper (Stone, 1954). Other studies have
directions for future research.
developed tax- onomies based on orientations to product
usage (Dardin & Reynolds, 1971), actual patronage and
shopping behav- ior (Stephenson & Willett, 1969),
Background and review of literature shopping-related AIO items (Moschis, 1976), shopping
enjoyment (Bellenger & Korgaonkar, 1980), and retail
Shopping research has long focused on the utilitarian attribute preferences (Bellenger, Robertson, & Greenberg,
aspects of the shopping experience, which has often been 1977; Dardin & Ashton, 1974).
characterized as task-related and rational (Batra & Ahtola, In a widely cited study, Tauber (1972) developed a num-
1991) and related closely to whether or not a product ac- ber of shopping motivations, with the basic premise that
quisition mission was accomplished (Babin, Darden, & shoppers are motivated by a variety of psychosocial needs
Griffin, 1994). However, traditional product acquisition ex- other than those strictly related to acquiring some product.
planations may not fully reflect the totality of the shopping These motives can be classified into personal (i.e., role
experience (Bloch & Richins, 1983). Because of this, the play- ing, diversion, self-gratification, learning about new
last several years have seen resurgent interest in shoppings trends, physical activity and sensory stimulation), and
hedonic aspects, particularly as researchers have social (i.e., social experiences, communication with others,
recognized the importance of its potential entertainment peer group attractions, status and authority, and pleasure of
and emotional worth (Babin et al., 1994; Langrehr, 1991; bargaining). Shopping thus occurs when a consumers need
Roy, 1994; Wakefield & Baker, 1998). for a par- ticular good is sufficient for allocating time and
Hedonic consumption has been defined as those facets of money to travel to a store to go shopping, or when a
behavior that relate to the multisensory, fantasy, and emo- consumer needs attention, wants to be with peers, desires
tive aspects of consumption (Hirschman & Holbrook, to meet people with similar interests, feels a need to
1982). This view suggests that consumption is driven by exercise, or simply has leisure time (Tauber, 1972, p. 48).
the fun a consumer has in using the product, and the Westbrook and Black (1985) linked Taubers (1972)
criteria for success are essentially aesthetic in nature framework to McGuires (1974) typology of 16 funda-
(Holbrook & Hirschman, 1982). Hedonic shopping motives mental human motivations, suggesting that shopping be-
are similar to the task orientation of utilitarian shopping havior arises for three fundamental reasons: to acquire a
motives, only the task is concerned with hedonic product, to acquire both a desired product and provide
fulfillment, such as expe- riencing fun, amusement, fantasy, satisfaction with non-product-related needs, or to primar-
and sensory stimulation (Babin et al., 1994). ily attain goals not related to product acquisition. These
Although the festive and ludic aspects of shopping fundamental shopping motives are captured in seven
have generally been studied infrequently (Sherry, 1990), dimen- sions of shopping motivation labeled, anticipated
hedo- nic aspects of shopping motivation have been utility, role enactment, negotiation, choice
uncovered in related phenomenological inquiry. For optimization, affiliation, power/authority, and
example, Christmas stimulation. While all motivations can be described as
containing both hedonic
 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795 79
grounds. The final sample included 33 men and 65 women,
and utilitarian elements, Westbrook and Black (1985) note ranging from 18 to 55 years of age. A variety of
that some are more utilitarian in nature while others are occupations and income levels were represented.
more hedonic in nature. We focus here on motivations that The interviewers were given a discussion guide and very
are primarily hedonic and non-product in nature. specific instructions as to how to conduct the interviews.
Interviewers provided a brief description of the goal of the
depth interview, and respondents were first asked to think
about shopping in general, in stores and/or malls (exclud-
Qualitative inquiry and initial scale development ing grocery shopping), and to describe reasons why they go
shopping, how they felt when shopping, and benefits they
We rely on the accepted paradigm for scale development received from shopping. The interviewers were instructed
provided by (Churchill, 1979) and augmented by others to probe the reasons, feelings, and benefits in depth by ask-
(e.g., Anderson & Gerbing, 1982; Bagozzi, 1980; Bentler ing extensive follow-up questions. All depth interviews
& Bonnet, 1980; Churchill, 1979; Gerbing & Anderson, were tape recorded and transcribed. Each respondents
1988; name and daytime telephone number were recorded for
Nunnally & Bernstein, 1994; Peter, 1981). Fig. 1 summa- research verifi- cation purposes, and each respondent was
rizes the scale development procedures employed here, and assured of his/her anonymity (i.e., no names were attached
the procedures are discussed in detail in subsequent to the interview notes). To ensure data quality, a random
sections. sample of respon- dents was contacted to validate the
interview, and no abnor- malities were noted.
Qualitative inquiry The interviews were read thoroughly many times by a
coding team (one of the authors and two graduate students).
Depth interviews were used to uncover the hedonic rea- Each member of the team (individually) identified and
sons people shop. This method was employed because it listed recurring themes in the data, using a categorizing
provides a deep understanding of a phenomenon from the process developed by Lincoln and Guba (1985). This
consumers perspective (Hudson & Ozanne, 1988). involved sort- ing themes into categories based on similar
Respon- dents were recruited on a referral basis98 characteristics. Then, the three members met to discuss the
undergraduate students were asked to provide the name and key themes (mo- tivations for shopping) and illustrative
demographic characteristics of a friend or family member quotes from the data. The goal at this point was to search
who would be willing to participate in a depth interview. for commonalities that allowed for the most accurate
Strict guidelines were given to ensure a diverse sample of representation of each domain and to develop conceptual
shoppers with re- gards to age, occupation, gender, income, definitions of the motivations. In addition, labels for each
and reasons for shopping. The students were instructed not motivation were constructed, and
to include other college students, and the initial list of
respondents was pre- screened by the authors to ensure that
the sample would include respondents with differing points
of view and back-
Fig. 1. Scale development process.
80 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

disagreements were resolved by discussion. Two market- interpersonal relationships. A significant amount of prior
ing faculty members familiar with the topic area evaluated re- search has uncovered social aspects of shopping
the motivations (accompanied by illustrative quotes from motivation. Stone (1954) first identified a personalizing
the data) and the corresponding conceptual definitions for shopper, one who seeks personal relationships while
content validity. shopping, whereas Moschis (1976) acknowledged a
psychosocializing shop- per. Tauber (1972) also
Qualitative results and domain definitions recognized that shoppers desire social interaction outside
the home, communicating with others having similar
Six broad categories of hedonic shopping motivations interests, and affiliating with reference groups. In addition,
emerged from the data. Appendix A illustrates sample com- Westbrook and Black (1985) identified affiliation as a
ments from informants for each of the six categories which shopping motivation, and Reynolds and Beatty (1999)
we have labeled as the following: adventure shopping, discuss social motivations for shopping.
so- cial shopping, gratification shopping, idea
shopping, role shopping, and value shopping. Each of Gratification shopping
these mo- tivations is briefly defined and discussed in light A third category is labeled gratification shopping,
of theo- retical explanations of human motivation and prior which involves shopping for stress relief, shopping to
research findings. alleviate a negative mood, and shopping as a special treat to
oneself. Several respondents admitted that they go
Adventure shopping shopping to re- lieve stress or to forget about their
The first category is labeled adventure shopping, problems. Other infor- mants view the shopping experience
which refers to shopping for stimulation, adventure, and the as a way to wind down, relax, improve a negative mood, or
feeling of being in another world. A significant number of just treat themselves.
respon- dents reported that they go shopping for the sheer Gratification shopping is grounded in McGuires (1974)
excite- ment and adventure of the shopping trip. These collection of tension-reduction theories of human
informants often described the shopping experience in motivation (e.g., Freud, 1933), which suggests that humans
terms of adven- ture, thrills, stimulation, excitement, and are moti- vated to act is such a way as to reduce tension,
entering a different universe of exciting sights, smells, and thereby main- taining inner equilibrium and returning the
sounds. self to a state of homeostasis. Babin et al. (1994)
Adventure shopping is grounded in stimulation theo- recognized the value of shopping as a self-gratifying,
ries (e.g., Berlyne, 1969) and expressive theories (e.g., escapist, and therapeutic activity, describing respondents
Huizinga, 1970; see also Sherry, 1990) of human moti- who view shopping as a pick-me-up and a lift when
vation as described by McGuire (1974). These theories they feel depressed. Tauber (1972) also identified the self-
are externally oriented, and stress the need for stimula- gratifying benefits of shop- ping, such that the process of
tion and self-expression through play and creativity among shopping to make the shopper feel better. Finally, shopping
human organisms. Adventure shopping is also similar to has been acknowledged in the literature as a form of
prior findings which show that shoppers often seek sensory emotion-focused coping in response to stressful events or
stimulation while shopping. For example, Tauber (1972), simply to get ones mind off a problem (Lee, Moschis, &
Westbrook and Black (1985) uncovered the personal shop- Mathur, 2001).
ping motive of sensory stimulation, Babin et al. (1994)
refer to adventurous aspects of shopping as a factor that Idea shopping
may produce hedonic shopping value, and Jarboe and A fourth category we label idea shopping, which refers
McDaniel (1987) identified shoppers (labeled browsers) to shopping to keep up with trends and new fashions, and
who enjoyed exploring and window shopping. to see new products and innovations. A significant number
of both females and males reported that they shop to keep
Social shopping up with the latest trends and fashions. Other informants
A second category is labeled social shopping, which describe shopping as a way to keep abreast with new
refers to the enjoyment of shopping with friends and fam- products and innovations that are available.
ily, socializing while shopping, and bonding with others Idea shopping is grounded in McGuires (1974)
while shopping. Respondents mentioned quite frequently collection of categorization theories, which collectively
that shopping is a way to spend time with friends and/or attempt to ex- plain the human need for structure, order,
family members. Some respondents stated that they just en- and knowledge, as well as objectification theories (e.g.,
joy socializing with others while shopping and that Festinger, 1954), which view the human as needing
shopping gives them a chance to bond with other shoppers. external guidelines and informa- tion in an attempt to make
Social shopping is grounded in McGuires (1974) col- sense of himself. This motivation corresponds with
lection of affiliation theories of human motivation (e.g., Taubers (1972) personal shopping motive of learning
Sorokin, 1950), which collectively focus on people being about new trends and keeping informed about the latest
altruistic, cohesive, and seeking acceptance and affection in trends in fashion, styling, or innovations. Some con-
sumers may enjoy browsing to obtain information as an
end in itself, not to make a particular purchase (Bloch,
Ridgway, & Sherrell, 1989). Bloch, Sherrell, and
Ridgway (1986)
 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795 81
Item generation
describe pleasure and recreationhaving fun and
experienc- ing positive affectas a motive for ongoing Based on the findings of the qualitative study, as well as
search (infor- mation gathering independent of a specific instruction from theory and ideas from prior research, items
purchase need or decision). Thus, for these consumers, were constructed to tap each of the six categories of
ongoing search rep- resents a leisure pursuit as an end goal shopping motivations. The initial item-generation process
(Punj & Staelin, 1983). produced 140 items: 29 items for adventure shopping, 28
items for gratification shopping, 25 items for social
Role shopping shopping, 11 items for role shopping, 21 items for value
A fifth category of shopping motivations is labeled role shopping, and 26 items for idea shopping.
shopping, which reflects the enjoyment that shoppers Several marketing faculty members then evaluated the
derive from shopping for others, the influence that this items for content and face validity. The faculty members
activity has on the shoppers feelings and moods, and the were given the conceptual definitions of the motivations,
excitement and intrinsic joy felt by shoppers when finding along with illustrative quotes from the data, and instructed
the perfect gift for others. Many respondents talked about to retain items based on their representation of the motiva-
the enjoyment they obtain from shopping for other people, tional domain and clarity of wording. Candidates for dele-
explaining that shopping for their friends and family is very tion were items that were not clear, not representative of the
important to them and that it makes them feel good. Some domain, or that were possibly open to misinterpretation
respondents described the positive feelings they get from (e.g., Babin et al., 1994). In addition, a substantial number
finding the perfect gift for someone. of re- dundant items were eliminated. The authors then
Role shopping is grounded in McGuires (1974) collec- reviewed the list of candidates for elimination and any
tion of identification theories of human motivation (e.g., inconsisten- cies were resolved by discussion. The resulting
Goffman, 1959), whereby people are motivated by the per- item pool contained 48 items: 9 items for adventure
ceived roles they may be playing at any given time. In shopping, 6 items for gratification shopping, 12 items for
essence, people seek ego enhancement to their self- social shopping, 4 items for role shopping, 8 items for
concepts through the addition of satisfying roles and value shopping, and 9 items for idea shopping. The item
acting out the roles responsibilities. This motive is pool was then submitted to a multi-sample scale
related to Taubers (1972) personal motive of role purification and validation process, which is described
playing, in which the pro- cess of shopping produces next.
positive effects for people who view it as part of their
social role. It also closely corre- sponds to Westbrook and
Blacks (1985) role enactment, which describes the drive to Scale purification
fulfill culturally prescribed roles regarding shopping. Babin
et al. (1994) explain how some consumers can view Substantive (e.g., breadth of theoretical content coverage
shopping as a duty, but enjoy the experience and obtain by an item) as well as empirical considerations were em-
hedonic value from the process. Further, other researchers ployed throughout the scale purification process (cf., Chin
have discussed how, for some consumers (especially & Todd, 1995). Scale purification is concerned with
women), shopping is an expression of love (cf. Miller, detailed item analyses, exploratory factor analyses,
1998; Otnes & McGrath, 2001). confirmatory factor analyses, and an initial assessment of
scale reliability, unidimensionality, and convergent and
Value shopping discriminant validity. Here, standard (e.g., Anderson &
The final category is labeled value shopping, which Gerbing, 1988; Churchill, 1979; Gerbing & Anderson,
refers to shopping for sales, looking for discounts, and 1988; Hair, Anderson, Tatham, & Black, 1998), as well as
hunt- ing for bargains. Many of our respondents talked emerging guidance (e.g., Peterson, 2000) in the literature is
about how they enjoyed hunting for bargains, looking for employed in item reduction and assessment of the resulting
sales, and finding discounts or low prices, almost as if factor structure.
shopping is a challenge to be conquered or a game to be A questionnaire was constructed that contained the 48
won. hedonic motivation items (7-point agreedisagree response
Value shopping is grounded in McGuires (1974) collec- format), interspersed throughout the questionnaire, as well
tion of assertion theories (e.g., McClelland, 1961), which as age, income and gender items. Respondents were in-
view the human as a competitive achiever, seeking success structed to think about shopping in stores and malls, and
and admiration, and striving to develop his potentials in or- not consider on-line/television shopping or convenience
der to enhance his self-esteem. Consumers may obtain he- formats such as grocery stores or drug stores. As used
donic benefits through bargain perceptions, which provide successfully in prior research (Bitner, Booms, & Tetreault,
increased sensory involvement and excitement (Babin et 1990; Gwinner, Gremler, & Bitner, 1998; Wallendorf &
al., 1994). Value shopping may also be related to the choice Arnould, 1991), marketing research students were recruited
optimization dimension identified by Westbrook and Black and trained as data collectors for the calibration sample.
(1985), given that finding a discount or bargain may lead to Respondents were contacted face-to-face or by
satisfaction from personal achievement. telephone, and subsequently
82 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

given a paper copy of the self-administered questionnaire. This procedure resulted in the deletion of 2 items for the
A total of 269 questionnaires were returned, and 3 adventure shopping dimension, leaving a remaining item
question- naires were judged unusable, leaving a final
sample size of n = 266. An inspection of the demographic pool of 34 items for further analysis.
variables re- vealed representation in all age and income
categories, with approximately 38% of the respondents Exploratory factor analysis
male and 62% fe-
male. Following item analysis, the 34 items were then sub-
jected to exploratory factor analysis with principal axis
Item analysis factoring and oblique rotation, with the scree test criterion
used to identify the number of factors to extract (Bearden
First, corrected item-total subscale correlations were ex- et al., 1989; Hair et al., 1998; Nunnally & Bernstein,
amined for each set of items representing a hedonic moti- 1994). A six-factor model was estimated, and items ex-
vation dimension. Items not having a corrected item-total hibiting low factor loadings (<.40), high cross-loadings
correlation above .50 were candidates for deletion (cf., (>.40), or low communalities (<.30) were candidates for
Tian, Bearden, & Hunter, 2001; Zaichowsky, 1985). After elimination (cf., Hair et al., 1998). After inspection of item
care- ful inspection of item content for domain content for domain representation, 9 items were deleted (2-
representation, 12 items having corrected item-total item cross-loadings >.40, 7-item factor loadings <.40). The
correlations of .50 and be- low were subsequently deleted remaining 25 items were submitted to further ex- ploratory
(7 items representing social shopping, 4 items representing factor analysis. Applying the same empirical and
value shopping, and 1 item representing idea shopping). substantive considerations in item trimming, 2 additional
Second, the correlations for items with their items were deleted (both item factor loadings <.40). A
hypothesized dimension were then compared with their final six-factor model was estimated with the remaining
correlations with the remaining dimensions (cf., Bearden, 23 items. The factor solution accounted for approximately
Netemeyer, & Teel, 1989; Tian et al., 2001). Items that did 67% of the total variance, and exhibited a KMO measure
not have statistically higher correlations (cf., Bruning & of sampling adequacy of .90. All communalities ranged
Kintz, 1977) with the dimensions to which they were from .41 to .77. Table 1 illustrates the 23-item factor
hypothesized to belong in comparison to other dimensions structure.
were subsequently deleted.

Table 1
Exploratory factor analysis results sample 1a

Items Adventure Value Role Idea Social Relaxation

shopping shopping shopping shopping shopping shopping
To me, shopping is an adventure .86 .03 .05 .02 .09 .05
I find shopping stimulating .67 .01 .06 .08 .13 .04
Shopping is a thrill to me .59 .05 .12 .06 .08 .19
Shopping makes me feel like I am in my own universe .57 .03 .08 .14 .03 .19
For the most part, I go shopping when there are sales .04 .86 .07 .03 .01 .11
I enjoy looking for discounts when I shop .05 .83 .06 .01 .01 .05
I enjoy hunting for bargains when I shop .05 .74 .01 .04 .02 .07
I go shopping to take advantage of sales .04 .73 .19 .17 .05 .04
I like shopping for others because when they feel good I feel good .15 .04 .83 .03 .01 .02
I feel good when I buy things for the special people in my life .05 .03 .81 .05 .05 .02
I enjoy shopping for my friends and family .02 .05 .80 .02 .10 .03
I enjoy shopping around to find the perfect gift for someone .07 .07 .57 .06 .04 .15
I go shopping to keep up with the trends .07 .04 .05 .87 .07 .01
I go shopping to keep up with the new fashions .01 .03 .10 .84 .08 .03
I go shopping to see what new products are available .06 .05 .01 .70 .05 .05
I go shopping to experience new things .17 .09 .09 .66 .05 .02
I go shopping with my friends or family to socialize .01 .01 .02 .04 .85 .01
I enjoy socializing with others when I shop .02 .01 .06 .02 .76 .06
To me, shopping with friends or family is a social occasion .06 .09 .02 .02 .71 .20
Shopping with others is a bonding experience .18 .07 .02 .01 .71 .05
When Im in a down mood, I go shopping to make me feel better .23 .04 .07 .05 .04 .69
To me, shopping is a way to relieve stress .13 .10 .07 .03 .15 .67
I go shopping when I want to treat myself to something special .08 .04 .11 .11 .14 .50
a
Pattern matrix shown. Principal axis factoring, oblique rotation. KMO measure of sampling adequacy = .90. Cumulative variance extracted = 67%.
 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795 83
only one underlying construct (Bollen, 1989; Gerbing &
Confirmatory factor analysis Anderson, 1988). As illustrated in Appendix B, coefficient
alpha estimates, ranging from .79 to .86, and the composite
reliability estimates, ranging from .83 to .89, are considered
The scale purification procedures rely on an iteration of
acceptable (Fornell & Larcker, 1981; Nunnally &
confirmatory factor analyses, with the goal to improve the
Bernstein, 1994). Item-to-total correlations, also appearing
congeneric measurement properties of the scale (Anderson
in Appen- dix B, range from 0.55 to 0.79, and all
& Gerbing, 1988; Bagozzi, 1980; Bearden et al., 1989;
variance ex-
Chin & Todd, 1995; Gerbing & Anderson, 1988;
tracted estimates (adventure shopping = .72; gratification
MacCallum, 1986). A 23-item, six-dimension confirmatory shopping = .71; social shopping = .73; value shopping
factor model was estimated using LISREL 8.1 (Jreskog & =
.71; role shopping = .67; idea shopping = .71) exceed the
Srbom, 1993), and inspection of model fit revealed
recommended .50 threshold (Fornell & Larcker, 1981).
indices that were
2
generally below acceptable thresholds ((215) = 604.07,
p = .000; GFI = .81; AGFI = .76; CFI = .89; NNFI = Convergent and discriminant validity
.87; standardized RMR = .064; RMSEA = .09). Item
squared multiple correlations (SMCs) ranged from .43 to Convergent validity can be assessed from the measure-
84, and an inspection of the modification indices (MIs) ment model by determining whether each indicators es-
revealed 4 items as candidates for removal, each account- timated maximum likelihood loading on the underlying
ing for three or more significant modification indices (MIs construct is significant (Anderson & Gerbing, 1988; see
ranging from 8.90 to 22.78). Each item was then inspected also Netemeyer, Johnston, & Burton, 1990; Peter, 1981).
for domain representativeness (cf., Nunnally & Bernstein, As illustrated in Appendix B, all confirmatory factor load-
1994). For example, the candidate item for removal for the ings exceed .66, and all are significant with t values ranging
social shopping dimension was, To me, shopping with from a low of 10.61 to a high of 17.31. Therefore, we have
friends or family is a social occasion, which tapped into evidence of convergent validity of our measures.
the same family and friends facet as the retained item, I Hedonic shopping motivations are conceptually related
go shopping with my friends or family to socialize. The constructs, yet are also expected to exhibit discriminant va-
three other candidate items exhibited similar facet charac- lidity (Westbrook & Black, 1985). As evidence of their
teristics, therefore, facet representation was ensured and the relationship, the inter-factor correlations between the five
four items were removed from further consideration. hedonic motivations, estimated by the phi coefficient,
A second confirmatory model was then estimated on the ranges from .23 to .72. Discrimination between the
remaining 19 items. Model fit was substantially improved: constructs is evident since the variance extracted estimates,
= 335.22, p = .000; GFI = .86; AGFI = .81;
2
(137) ranging from
CFI = .92; NNFI = .90; standardized RMR = .049; .67 to .73, exceed all squared phi correlations, ranging from
RMSEA = .08. Item SMCs ranged from .44 to .87, and an
examination of modification indices revealed three signifi- .05 to .52 between the constructs (Bearden et al., 1989;
cant indices. The item having the largest significant modifi- Fornell & Larcker, 1981; Netemeyer et al., 1990).
cation index (10.32) was, I go shopping to experience new Therefore, we have evidence of discriminant validity.
things, which was judged to not represent the idea shop-
ping facet of new products as well as the retained item,
I go shopping to see what new products are available. Scale validation
Therefore, this item was removed from further analysis.
A final confirmatory model was then estimated on the
remaining 18 items. The model exhibited respectable fit: The purpose of scale validation activities is fourfold.
= 289.42, p = .000; GFI = .87; AGFI = .82; CFI = First, it is desirable to replicate the confirmatory factor
2
(120)
structure on an independent sample, thereby reducing error
.93; NNFI = .91; standardized RMR = .049; RMSEA =
.079. All modification indices were predominantly low, al- due to capitalization on chance (Chin & Todd, 1995;
though two were marginally significant, and item SMCs MacCallum, Roznowski, & Necowitz, 1992). Not only
ranged from 0.44 to 0.87. Since the final 18 items parsimo- should the model replicate, but we must also show the
niously represent the six hedonic dimensions, and since extent to which our measurement model is stable across
each item taps into a unique facet of each hedonic independent samples. Second, the hedonic constructs are
dimension and thus provides good domain representation, then correlated with the- oretically related constructs,
no further items were removed. See Appendix B for thereby establishing evidence of nomological validity.
complete item measure- ment properties of the final six- Third, to demonstrate the useful- ness of the scale, some
dimension, 18-item scale. degree of predictive validity of the hedonic measures must
be shown. Finally, we use the data to cluster respondents
Unidimensionality and reliability into meaningful shopper segments, thereby providing
additional practical utility of the scale for retailers.
A two-part questionnaire was constructed that contained
Given these results, we have evidence that the mea- the 18 hedonic motivation items in the first part, and a vari-
sures are unidimensional, with each item reflecting one and ety of variables used for nomological and predictive
validity
84 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

tests, and age, income and gender items in the second part. sional measures (e.g., low and/or insignificant modification
The data collection procedures employed for the calibration indices; see Gerbing & Anderson, 1988). As appearing in
sample were replicated here for the validation sample, only Appendix B, reliability of the subscales is acceptable as co-
students were instructed to administer the two parts of the efficient alpha estimates range from .77 to .87 (Nunnally &
questionnaire approximately 2 weeks apart. Administering Bernstein, 1994). Composite reliability estimates (Fornell
the hedonic motivation items separately reduces the com- & Larcker, 1981) range from .80 to .92, corrected item-to-
mon methods bias that might explain correlations between total correlations range from .50 to .83, and all variance
the hedonic constructs and other related variables. Both the extracted
sampling procedures and instructions for respondents em- estimates range from .58 to .78 (adventure shopping = .78,
ployed for the calibration sample were also employed here. gratification shopping = .58, social shopping = .72, value
A total of 253 completed surveys were returned, and two shopping = .76, role shopping = .71, idea shopping = .
75). Convergent validity is evident in that all confirmatory
questionnaires were judged unusable, leaving a final factor
sample loadings exceed .61 and are significant (t values range from
size of n = 251. Respondent names and contact informa- 9.32 to 17.89).
tion were recorded, and a random subsample of 25 respon- Discriminant validity was again tested by comparing the
dents was contacted to verify the research procedures they variance extracted estimates with the squared phi correla-
followed. No issues or abnormalities were noted. The tions between the hedonic constructs (Fornell & Larcker,
demo- graphic profile of the validation sample was highly 1981). Two squared phi correlations, gratification/adventure
consis- tent with that of the calibration sample: all age and shopping (22 = .76, = .87) and gratification/social
income categories were represented with approximately shopping ( = .64, = .80) exceeded the variance ex-
32% of the respondents male and 68% female. tracted estimate for gratification shopping of .58, but not
for
social (.72) or adventure shopping (.78). Therefore, addi-
Factor structure stability tional tests for discriminant validity were conducted. First,
chi-square difference tests were performed among several
A measurement model was then estimated using the 18 models that fixed the relationship between the suspect con-
items developed in scale purification. The results indicated structs to one (Anderson & Gerbing, 1988; see also
good fit (2(120)= 254.15, p = .000; GFI = .88; AGFI = Bearden et al., 1989): gratification/adventure shopping
.83; CFI = .94; NNFI = .92; standardized RMR = .048; fixed to unity
RMSEA = .073). Several modification indices were signif- (2(121)= 419.71, p = .000; RMSEA = .105) and grat-
icant but predominantly low (ranging from 8.24 to 17.65),
2
and item squared multiple correlations ranged from 0.37 ification/social shopping fixed to unity (121)( =
444.18, p = .000; RMSEA = .109). The fit for both
to 0.91. No modifications were made to the measurement 2
models was significantly worse (6y with 1 degree of
model because (a) no theoretical or conceptual basis justi- freedom) when
fies making further modifications, (b) the model fits the compared to the fit of the theoretically specified six-factor
data well and replicates across independent samples. model. Therefore, this test shows support for the discrim-
To assess the factorial stability of the hedonic motiva- inant validity between the three hedonic motivations. A
tions, a multi-group analysis procedure was performed in second test for discriminant validity was then performed
LISREL 8.1 that allows for the independent estimation of that examines the confidence interval around the correlation
factor loadings, factor correlations, and error variances on between the suspect factors. If the correlation plus or mi-
the two samples (Byrne, 1998; Jreskog & Srbom, 1993; nus 2 standard errors does not include the value 1.0, then
Steenkamp & Baumgartner, 1998). Six multi-group tests evidence of discriminant validity is shown (Anderson &
revealed substantial measurement equality across the cali- Gerbing, 1988). Neither confidence interval for the phi es-
bration and validation samples. When compared to the timates between gratification/adventure shopping (.81.93)
base- line model 2
= 614.81), successive models and gratification/social shopping (.72.88) includes the
(263) (
showed no value of 1.0. In summary, since two of three statistical
significant change in chi-square: equal factor loadings only tests provide evidence of discriminant validity, and since
2
2 discriminant validity was previously shown between all
( (273) = 623.13); equal factor correlations only ((286) = constructs on an independent sample (calibration sample),
2
625.89); equal error variances only ((286) = 622.40); we have sufficient evidence of discriminant validity.
equal
factor loadings and factor correlations (2(268) = 621.04);
and equal factor loadings, factor correlations, and error Nomological validity
variances (2(291) = 628.59). Therefore, we have evidence
of the factorial stability of the hedonic motivations scale The importance of establishing nomological validity has
across independent samples. been well documented in the literature (e.g., Babin et al.,
1994; Cronbach & Meehl, 1955; Netemeyer, Durvasula,
Reliability and validity & Lichtenstein, 1991). Therefore, the six hedonic shop-
Unidimensionality, reliability, convergent and discrimi- ping constructs were investigated within a larger nomolog-
nant validity were then evaluated. Given the results of the ical network of theoretically related constructs, including
model estimation, we again have evidence of unidimen-
flow (Csikszentmihalyi, 1990), time distortion and
aesthetic
 M
.J.
Ar
no
ld,
K.
Table 2 E.
Nomological validity assessmenta R
Gratification Social Role Value Idea Flow Time Aesthetic Product Non-generosity Personal ey
Adventure
shopping shopping shopping shopping shopping distortion appeal innovativeness shopper no
shopping ld
Adventure shopping 1.00 s/
Jo
Gratification shopping .71 1.00 ur
Social shopping .63 .58 1.00 na
Role shopping .38 .46 .42 1.00 l
Value shopping .34 .42 .32 .37 1.00 of
R
Idea shopping .48 .52 .46 .31 .17 1.00
et
Flow .55 .46 .38 .32 .24 .33 1.00
ai
Time distortion .62 .59 .49 .39 .31 .36 .60 1.00 li
Aesthetic appeal .45 .39 .40 .29 .30 .33 .33 .42 1.00 ng
Product innovativeness .26 .28 .17 .20 .09b .42 .15 .28 .28 1.00 79
Non-generosity .13 .17 .30 .43 .31 .07b .15 .18 .21 .11b 1.00 (2
Personal shopper .26 .21 .28 .05b .19 .16 .11b .17 .19 .07b .03b 1.00 00
a
All correlations significant at p< .05 (two-tailed) unless otherwise noted.
b
Correlation not significant at p> .05 (two-tailed).

85
86 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

appeal (Bloch, Ridgway, & Dawson, 1994), innovativeness H2c. The correlation between time distortion and adventure
(Oliver & Bearden, 1985), non-generosity (Belk, 1984), shopping will be significantly higher than the correlations
and personalizing shopper attitudes (Hawes & Lumpkin, between time distortion and role, value, social, and idea
1984). Each of these is briefly discussed in order, and the shopping.
correlation
estimates calculated from the validation sample (n = 251)
appear in Table 2. H2d. The correlation between time distortion and gratifi-
cation shopping will be significantly higher than the corre-
Flow lations between time distortion and role, value, social, and
Flow is a cognitive state that has been characterized idea shopping.
as an optimal experience that is intrinsically enjoyable
(Csikszentmihalyi, 1990). Consumers experiencing flow To measure time distortion, we employed three items
are deeply involved in the focal activity where time may ( = .92, 7-point format ranging from strongly disagree
to strongly agree) used in prior research (Bloch et al.,
seem to stand still (Hoffman & Novak, 1996). Since 1994). An example of one such item employed is, I lose
adventure shopping captures the experiential and fantasy track of time when Im in a store or at the mall. As seen in
aspects of shopping, we should expect flow to positively Table 2, the correlations between time distortion and adven-
correlate with adventure shopping, and be more strongly ture shopping (r = .62) and gratification shopping (r = .
correlated with adventure shopping than with the other 59) are both significantly higher (p< .05) than the
hedonic motivations. Therefore: correlations between time distortion and the remaining
hedonic motiva-
tions. Therefore, H2a, H2b, H2c, and H2d are supported.
H1a. The correlation between flow and adventure shopping
will be positive.
Aesthetic appeal
Aesthetic appeal refers to an appreciation of the physical
H1b. The correlation between flow and adventure shopping design or appearance of the mall habitat, and is based on
will be significantly higher than the correlations with other the premise that some consumers notice and enjoy the
hedonic motivations. physical elements of the retail environment (Bloch et al.,
1994). Re- search has shown the importance of the retail
atmosphere in influencing a wide variety of emotions and
Flow was measured with a single item used by Novak, behaviors re- lated to shopping (e.g., Donovan & Rossiter,
Hoffman, and Yung (2000) and adapted for the present 1982). Thus, we should expect that aesthetic appeal is
study: Most of the time I go shopping at malls or stores I likely to be posi- tively correlated with all of the hedonic
feel that I am in flow (7-point format ranging from motivations studied here. Therefore:
strongly disagree to strongly agree). The item was preceded
with a description of flow also used by Novak et al. (2000). H3a. Aesthetic appeal will correlate positively with all six
The correlation estimate reported in Table 2 between flow hedonic motivations.
and adventure shopping of .55 is significantly higher (p < .
05) than correlations with other hedonic motivations.
However, role shopping, value shopping, and idea shop-
Therefore, H1a and H1b are supported.
ping address motivations that are product-oriented (i.e.,
find- ing the perfect gift, shopping for others, hunting for
Time distortion bargains, finding new fashions, etc.), as opposed to the
Time distortion describes a psychological state where non-product motives of social interaction (social shopping),
shoppers become relatively isolated from the cues regard- stimulation and fantasy (adventure shopping), or stress
ing the passage of time (Bloch et al., 1994). Hours pass relief and re- laxation (gratification shopping). Given this
without notice if the consumption state is sufficiently pleas- reasoning, we should expect that the correlations between
ant (Kowinsky, 1985). Therefore, it is likely that shoppers aesthetic appeal and role shopping, value shopping, and
motivated by adventure shopping will also experience time idea shopping will be lower than the correlations with
distortion. Additionally, since gratification shopping mo- adventure shopping, grat- ification shopping and social
tives are oriented to relaxation, stress relief, and improved shopping.
mood states, it is reasonable to expect a strong correlation
with gratification shopping as well. Therefore:
H3b. The correlations between aesthetic appeal and role
shopping, value shopping and idea shopping, will be signif-
H2a. The correlation between time distortion and icantly lower than the correlations with the remaining
adventure shopping will be positive. hedo- nic motivations.

To measure aesthetic appeal, we employed three items

H2b. The correlation between time distortion and gratifica- ( = .90, 7-point format ranging from strongly disagree
tion shopping will be positive. to strongly agree) used in prior research (Bloch et al.,
 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795 87
tively with role shopping, and be significantly lower than
1994). An example of one such item employed is, I notice other correlations. Therefore:
colors or textures on the interior of retail stores or malls. H5a. The correlation between non-generosity and role
As seen in Table 2, the correlations between aesthetic ap- shopping will be negative.
peal and all six hedonic motivations are positive and signif-
icant. Therefore, H3a is supported. Further, the correlations H5b. The correlation between non-generosity and role
between aesthetic appeal and role shopping (r = .29), shopping will be significantly lower than the correlations
value shopping (r = .30), and idea shopping (r = .33) are
the lowest reported correlations. However, only the between non-generosity and other hedonic motivations.
correlation
between aesthetic appeal and adventure shopping is statisti-
To measure non-generosity, we employed five items
cally higher than the correlations with role, value, and idea ( = .69, 7-point format ranging from strongly disagree
shopping. Therefore, H3b is partially supported. to strongly agree) used in prior research (Belk, 1984).
An example of one such item employed is, I dont like
Product innovativeness to lend things, even to good friends. As seen in Table 2,
the correlation between non-generosity and role shopping
Innovativeness has been defined as the degree to which
(r = .43) is negative and significantly lower (p < .05)
an individual makes innovative decisions independent of than the correlations between non-generosity and the re-
infor- mation provided by others (Midgley, 1977). maining hedonic motivations. Therefore, H5a and H5b are
Innovativeness is viewed as a generalized personality trait, supported.
closely related to a number of dimensions of human
personality (Midgley & Dowling, 1978). Product
innovativeness is similar to the generalized trait Personalizing shopper
innovativeness, but relates to a persons de- sire to be Personalizing shopping reflects a consumers desire to
among the first to try new products (Oliver & Bearden, shop at stores where she/he is known (Dardin & Reynolds,
1985). Given our domain definition of idea shop- ping, we 1971). Personalizing shoppers prefer to individualize the
should expect a strong and positive relationship between shopping trip by seeking social relationships with other
idea shopping and product innovativeness. Further, this customers, and retail and service salespeople (Hawes &
relationship should be stronger than other reported re- Lumpkin, 1984; Stone, 1954). Since personalizing shop-
lationships. Therefore, per contains a strong social element, we should expect
a positive relationship with social shopping motives, and
the correlation between social shopping and personalizing
H4a. The correlation between innovativeness and idea shopper should be stronger than other reported correlations.
shopping will be positive. Therefore,

H6a. The correlation between personalizing shopper and

H4b. The correlation between innovativeness and idea
social shopping will be positive.
shopping will be significantly higher than the correlations
between innovativeness and other hedonic motivations.
H6b. The correlation between personalizing shopper and
social shopping will be significantly higher than the cor-
To measure product innovativeness, we employed two relations between personal shopping and other hedonic
items (r = .72, 7-point format ranging from strongly dis- motivations.
agree to strongly agree) used in prior research (Oliver &
Bearden, 1985). An example of one such item employed is,
I like to buy new and different things. As seen in Table 2, Role shopping is defined here as the motivation to shop
the correlation between innovativeness and idea shopping for others, and is based on the fundamental human motiva-
(r = .42) is positive and significantly higher (p< .05) than tion of acting out and fulfilling ones roles as people seek
the correlations between innovativeness and the remaining ego enhancement to their self-concepts. Since role
hedonic motivations. Therefore, H4a and H4b are shopping is product-oriented (e.g., gift shopping) and is
supported. inherently inner-directed (ego enhancement, empathy), the
social in- teraction aspects would seem less important than
Non-generosity in other shopping motivations. Given this reasoning, we
The conceptual domain for non-generosity includes the should see a weaker correlation with personalizing shopper
unwillingness to give or share possessions with others, a attitudes than with other hedonic motivations. Therefore:
reluctance to lend or donate possessions to others, and neg-
ative attitudes toward charity (Belk, 1985). Non-generosity H6c. The correlation between personalizing shopper and
has been found to be negatively related to happiness (Belk, role shopping will be significantly lower than the cor-
1984). Importantly, evidence suggests that non-generosity relations between personal shopping and other hedonic
may be based on egoist self-interest, with those more non- motivations.
generous people believing themselves to be unworthy to
give or receive (Belk, 1985; Neisser, 1973). Hence, we
should expect that non-generosity should correlate nega-
88 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

To measure personalizing shopper, we employed three Correlation analysis revealed our measure of browsing
items ( = .89, 7-point format ranging from strongly
disagree to strongly agree) used in prior research behavior correlated positively (p < .05) with the six hedo-
(Hawes nic motivations: adventure shopping (r = .43), gratification
& Lumpkin, 1984; Dardin & Reynolds, 1971). An shopping (r = .49), social shopping (r = .48), role shop-
example of one such item employed is, I like to shop ping (r = .28), value shopping (r = .34), and idea
where people know me. As seen in Table 2, the shopping (r = .38). Therefore, H7 is supported.
correlation between personalizingness and social Shopper segments
shopping (r = .28) is positive. Therefore, H6a is
supported. However, this correlation is only significantly
higher (p < .05) than the correlation To assess the practical utility of the hedonic shopping
between personalizingness and role shopping. Therefore, motivations scale, a shopper taxonomy was developed. A
H6b is partially supported. Finally, the correlation between two-step clustering procedure was employed for partition-
personalizingness and role shopping is not significant, and ing activities, using both hierarchical and non-hierarchical
is significantly lower (p < .05) than all other correlations. methods on the validation sample (n = 251). The first step
Therefore, H6c is supported. in partitioning was to submit the summed scales
representing
the six hedonic dimensions to hierarchical cluster analysis
Predictive validity (Wards method, squared Euclidian distances; see Milligan,
1980; Punj & Stewart, 1983). Given that prior research on
Predictive validity is defined as the ability of a shopper taxonomies suggests a variety of the number of
measuring instrument to estimate some criterion behavior opti- mal clusters (e.g., Bellenger & Korgaonkar, 1980;
that is exter- nal to the measuring instrument itself, and is Bellenger et al., 1977; Bloch et al., 1994; Westbrook &
shown by the correlation between the instrument and the Black, 1985), a range of cluster solutions (36) was tested,
criterion variable (Nunnally & Bernstein, 1994). To assess and an ex- amination of the dendrogram as well as the
predictive validity of the six hedonic motivations, a agglomeration schedule produced support for a five-cluster
measure of browsing be- havior was employed as the solution (see Hair et al., 1998; Milligan & Cooper, 1985 for
criterion variable. Browsing is understood as the ongoing a discussion of stopping rules).
search behavior in a retail envi- ronment for informational The next step consisted of employing k-means cluster
and/or recreational purposes and without immediate analysis using the hierarchical cluster centers as initial
intentions to purchase (Bloch & Richins, 1983; Bloch et al., seeds (Hair et al., 1998; Milligan, 1980). Table 3 illustrates
1989). the cluster means of the summed motivation scales
The extent and nature of browsing behavior is likely to (columns la- beled specified seeds), which was the final
be predicted by all six hedonic motivations held by shop- assignment of
pers. Prior research suggests that browsing behavior is cases to clusters, resulting in five clusters of n1 = 28, n2 =
recre- ationally motivated (Bloch et al., 1989; Jarboe & 34, n3 = 55, n4 = 64, and n5 = 55. As shown, ANOVA
models indicate significant mean differences across the five
McDaniel, 1987), related to the need for sensory clusters (F values ranging from 112.67 to 50.08). Further,
stimulation and di- version (Bloch et al., 1989), and also Tukey HSD post hoc tests illustrate differences between
provides a shopping context for social interaction (Bloch et spe- cific cluster means.
al., 1994; Jarboe & McDaniel, 1987). Further, browsing
provides shoppers with gratification and pleasure, as Cluster validation
consumers can vicari- ously buy desirable products Two validation procedures were performed. First, a k-
(Hirschman, 1980), and the informational dimension of means cluster analysis was performed using the six
browsing provides shoppers the benefits of seeing new summed shopping motivation scales, this time with ran-
products, fashions, and trends (Bloch et al., 1989), as well dom initial seeds (Hair et al., 1998; Punj & Stewart, 1983).
as providing a venue to hunt for bar- gains and shop for The results, appearing in Table 3 (random seeds columns),
others. confirm the stability of the five-cluster solution identi-
To measure browsing behavior, we employed three items fied during the partitioning activities. The cluster sizes as
( = .90) used in prior research (Jones, Reynolds, Weun, &
Beatty, in press). An example of one of the items is, How well as all cluster centroids are nearly identical across the
often do you visit stores just to look at new products, with specified-seeds and the random-seeds cluster models. The
no intention of making a purchase? Each item was mea- second step assesses predictive validity by investigating
sured on a 7-point scale ranging from Never to Very whether variables that are theoretically related to the hedo-
Frequently. Since predictive validity is shown by a sig- nic shopping motivations do indeed differ across clusters.
nificant correlation between two focal constructs (Nunnally Table 3 illustrates the results of univariate ANOVAs of the
& Bernstein, 1994), we hypothesize that all six motivation summed scales across the five clusters, along with Tukey
constructs will be correlated with the measure of browsing HSD post hoc tests.
behavior. Therefore: Given the information in Table 3, we provide an interpre-
tation of the clusters: Cluster 1, labeled the Minimalists,
H7. All six hedonic motivations will correlate positively is
with browsing behavior.
Table 3
Results of non-hierarchical cluster analysis and validationa
Shopping motivation Cluster means

Cluster 1, Minimalists Cluster 2, Gatherers Cluster 3, Providers Cluster 4, Enthusiasts Cluster 5, Traditionalists Specified seeds model

Specified Random Specified Random Specified Random Specified Random Specified Random F value Significant F
seedsb seedsc seeds seeds seeds seeds seeds seeds seeds seeds M
Hedonic motivation .J.
Ar
Adventure shopping 4.14e 4.23 5.38e 5.98 6.60 6.61 14.42 14.63 9.80 11.36 106.94 .00
no
Gratification shopping 5.71 5.71 7.82 8.45 9.69 9.73 16.84 17.00 13.36 14.90 112.67 .00
ld,
Role shopping 7.07 7.29 12.15e 13.12 17.16f 16.50 17.98f 18.33 13.05e 14.29 83.66 .00 K.
Value shopping 11.46 10.87 6.65 7.88 17.15e 17.95 16.94e 17.46 14.73 14.80 63.96 .00 E.
Social shopping 4.64 4.77 7.74e 8.00 8.51e 8.93 15.08 16.29 10.73 11.02 68.26 .00 R
Idea shopping 4.46 4.65 10.44 10.45 8.31 9.14 14.39 14.79 12.29 12.36 50.08 .00 ey
no
Demographics ld
Cluster size 28 31 34 42 55 56 64 42 55 59 s/
Percentage of respondentsd 12 13 15 18 23 24 27 20 23 25 Jo
Male (%) 57 70 17 10 42 ur
Femaleg (%) 43 30 83 90 58 na
Ageg l
<25 years old (%) 21 47 22 43 43 of
2549 years old (%) 54 41 63 36 46 R
et
Related constructs ai
Flow 1.56h 2.18h 2.46h,i 4.08 2.48i 19.74 .00 li
Time distortion 6.36 9.35h 10.65h 16.51 12.31 29.87 .00 ng
Aesthetic appeal 8.32h 10.09h,i 11.66i,j 14.95 12.70j 15.13 .00 79
Innovativeness 7.64h 8.82h,i 8.44i 9.68j 9.48j 6.27 .00 (2
Non-generosity 22.48h,i 21.55h 26.05j 26.90j 23.09j 8.40 .00 00
Personalizing shopper 7.43h 7.97h 9.04 10.85i 10.05j 3.95 .00
Browsing behavior 5.96 8.38h 9.33h 13.31 11.78 20.37 .00
a
Cluster means shown in cells.
b
The k-means cluster analysis performed with initial seeds specified from hierarchical cluster solution (Wards method, squared Euclidian distance).
c
The k-means cluster analysis performed with random seeds. Post hoc tests not show for clarity.
d
Percentage of respondents calculated on validation sample size of n = 236 following deletion of cases with missing values.
e,f
Not significantly different within hedonic factor (rows) at p< .05.
g
Significant differences between males and females (2 = 50.9, p< .01) and age groups (2 = 15.4, p< .05).
h,i,j
Not significantly different within rows at p< .05, post hoc HSD test.

89
90 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

composed of a majority of males and largely middle-aged, Faber, 1989), would also appear to be another application
and scores lower on all hedonic motivations with the of the scale for retailers and researchers alike.
excep- tion of value shopping. Cluster 2, labeled the Prior research has called for the investigation into which
Gatherers, is composed largely of younger males who motivations are likely to vary across retail shopping for-
score higher on idea and role shopping, and exhibit the mats, and temporal variation in shopping motivations
lowest level of value shopping. The Gatherers appear to be across a number of shopping occasions (Westbrook &
motivated by the hedonic aspects of gathering information Black, 1985). Important strategic considerations arise when
on new prod- ucts and trends, perhaps in anticipation of considering that different hedonic motivations may
future purchases. Cluster 3, labeled the Providers, is dominate in different retail formats. In addition, the
composed largely of middle-aged females who score highly importance of different moti- vations may vary with
on role and value shopping, and score the lowest in non- regards to the degree of the shoppers product involvement
generosity. Cluster 4, the Enthusiasts, is composed and the particular shopping situation. How the motivations
overwhelmingly of younger females and scores highly on vary with regards to gender and the specific shopping
all hedonic motivations. Clus- ter 5, the Traditionalists, is context is also an interesting question. In this study, females
slightly more females than males and younger to middle- scored higher on the hedonic motiva- tion subscales than do
aged, and scores moderately high on most hedonic males. This finding is consistent with Reynolds and Beatty
dimensions. (1999), however, other research has found that browsing
and ongoing search for computers and related products
were higher among men (Bloch et al., 1986; Otnes &
Discussion and implications McGrath, 2001). This is an area that requires fur- ther
research.
Implications for research
Implications for retailers
The hedonic shopping motivation scale captures a wide
variety of hedonic reasons people go shopping, and has A number of implications for retailers are apparent as
a broad variety of applications to retail research. First, well. First, knowledge of distinct shopper segments is use-
the scale can be employed in research investigating the ful for retailers in constructing marketing communication
interrelationships between hedonic motivations, in-store strategy and designing appealing store environments. For
ex- periences, and shopping outcomes (e.g., satisfaction). ex- ample, advertising in many cases may need to be
Prior research suggests that shopping motives drive the designed to attract shoppers who are motivated for different
behavior that brings shoppers into the marketplace, but the hedonic reasons. This could be accomplished by focusing
emotions experienced in the store affect preference and on the ex- periential aspects of the store environment,
choice behav- iors (Dawson, Bloch, & Ridgway, 1990). positioning the shopping experience as an adventure or a
Shoppers driven by a larger set of hedonic motivations may chance to visit with friends.
pay attention to a larger set of retail attributes (e.g., Store atmospherics can be tailored to certain shopper
merchandise displays, in-store promotions), and thereby seg- ments as well. If a retailer finds a large segment of En-
have a larger number of inputs in the decision-making thusiasts or Traditionalists among its regular customers, it
process (Dawson et al., 1990). could consider ways to facilitate the social experience its
Alternatively, intense shopping motives may create a customers can have. We see evidence of these
strong goal-attainment drive for consumers (Dawson et al., considerations in todays marketplace, particularly with the
1990). Hence in a manner similar to product involvement bookstorecaf concept that has become so popular (e.g.,
(Oliver, 1997), strong (vs. weak) motivations may magnify Barnes and No- ble). Conversely, if a store finds a large
the experience in the mind of the shopper. This could have segment of Gatherers among its target market, it may
the effect, like involvement, of making in-store evaluations consider providing more in- formative, hands-on displays
and affective responses more intense, either positive or that are increasingly popular in specialty stores (e.g., the
negative. Discovery Store). Alternatively, a retailer who finds that a
Another application centers on relating the type and large proportion of its customers are Providers may design
inten- sity of hedonic motivations to specific shopping an in-store experience character- ized by convenience and
behaviors, such as impulse purchasing and compulsive even an emphasis on web-based kiosks for ordering
consumption. Impulse purchasing has strategic value for products on-line.
retailers (Jones et al., in press), and is defined as the degree Aside from the shopper segmentation, retailers could use
to which an in- dividual is likely to make unintended, the scale to investigate the direction and strength of
immediate, and unre- flective purchases. Impulse hedonic motivations among their current customers. For
purchasing is also closely linked to hedonic consumption example, a hedonic motivational profile could be
and sensory stimulation (Beatty & Ferrell, 1998; Rook, constructed of a stores current customer base, thereby
1987). Compulsive consumption, which in a retail setting providing the retailer with additional knowledge of
describes shoppers who carry out shop- ping activities in intensity and types of hedonic motivations influencing its
pursuit of hedonic fulfillment (OGuinn & customers. Further, the retailer would be well positioned to
assess motivational strength of
 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795 91
or service, and to have fun as well. Therefore, measuring
different shopper groups. The retailer could focus on one only hedonic motivations may be missing part of the story.
or all of the hedonic motivations in investigations of mean This is both a limitation as well as a direction for further
levels of motivation across customer groups based on age, research. In addition, the current study only addressed
gender, race/ethnicity, and even profitability. shop- ping motivations in a single retail channelshopping
Retailers may wish to employ the scale to assess the in reg- ular stores. Further testing in other retail channels,
effects of different hedonic motivations on important shop- namely, on-line and catalog (e.g., Mathwick, Malhotra, &
ping outcomes. Retail research suggests a direct link be- Rigdon, 2001), may be warranted.
tween shopping motivations and outcomes such as retail While our qualitative investigation revealed a rich array
satisfaction, loyalty, and assessments of hedonic and utili- of hedonic shopping motivations, interesting was what our
tarian shopping value (Babin et al., 1994). For example, are qualitative investigation did not show, such as motives to
shoppers who are strongly motivated by hedonic dimen- interact with salespeople or derive feelings of power and
sions more likely to be satisfied, repatronize the retailer, authority from being waited on (e.g., Westbrook & Black,
and engage in behaviors such as talking positively about 1985). In hindsight, this may not be too surprising given
their experiences? Prior research suggests that satisfaction the increasingly poor levels of service consumers receive in
is a direct indicator of a shoppers motivational strength todays retail and service marketplace (Brady, 2000). Fur-
(Westbrook & Black, 1985), thereby suggesting that motive ther, shoppers may not easily admit to these motives, or
strength is directly and positively associated with aspects may not raise these motives when thinking about the
of preference and satisfaction (Dawson et al., 1990). There- hedonic motivations to shop. This clearly is another area
fore, retailers could potentially have another tool to manage for future research.
customer satisfaction and loyalty. Finally, it is important to note that the purified scale cap-
tures broad dimensions of shopping motivation, yet is also
parsimonious. Future research could investigate further a
number of areas that were uncovered here in the qualitative
Limitations and directions for future research investigation, such as value shopping, which relates to the
hunting for bargains, and winning the game of shop-
As with any scale development research, one must use ping (Sherry, 1990). Research could also investigate the
caution with the application of the scale to other shopping role of flow in shoppers evaluation of their shopping
contexts. Specifically, while we have provided evidence experi- ences along hedonic or utilitarian dimensions
that the scale replicates well across independent samples, (Babin et al., 1994), as well as a qualitative validation of
further evidence of generalizability is needed. Second, as the shopper seg- ments, and correlating each segment with
with any factor analysis, a certain amount of subjectivity is actual shopping behaviors. Finally, future research,
neces- sary in identifying and labeling factors. Third, the possibly experimental, could investigate hedonic
scale was developed only to address hedonic, or non- motivations after controlling for in-store mood states, as
product reasons people shop. Westbrook and Black (1985) well as the correspondence of the hedonic scores in spousal
and others have recognized that shoppers often have dyads.
utilitarian motives as well, in many cases going shopping to
acquire some product

Appendix A. Sample of respondent comments from qualitative study

Motivation Illustrative comments

Adventure
shopping It gets me all excited! So its kind of like exploring, only in a shoppers world? In a shoppers world.
Right ... Whats out there since Id last been there? Okay. Well, thats a type of adventure. Mm hm. Its
an adventure for me.
I enjoy shopping. It brings me great excitement and sometimes suspense as to what I am going to find.
Oh yeah, I always think about other things when I am shopping. Being in a different place helps me get
away from my everyday life. With clothes, I visualize where I would wear things. I think about
where I will wear things and imagine how everyone will think I am really pretty.
When I go to a store or mall, I am kind of in my own little shopping world. I dont try to think of
anything but what I like, what would I look good in, and what is eye catching enough for me to spend
my money on.
Well, I shop because it gives me a chance to spend time with my friends and family. I do not always go
shopping to buy things. I do a lot of shopping with my mom and my aunt and I feel its an excuse to
spend the day together.
Social
I go with my mom, we use shopping as a bonding time together.
shopping
92 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

Appendix A. (Continued )

Motivation Illustrative comments

The experience itself is kind of secondary unless I am out shopping with a friend or something. If you
are with a friend then it is more like a social atmosphere where you are going shopping as just an
excuse to hang out together and if you see something you want then you buy it.
Shopping for fun to me means more of socialization, if I had to pick one. But mostly socialization
with other people. I dont like to shop for fun by myself; I have to go with someone else for it to be any
fun.
I usually dont shop alone, so I socialize with the people Im with, and sometimes I run into people I
know while Im shopping.
Gratification
shopping I love to go shopping. It is my biggest stress reliever.
I like to go shopping when Im stressed; to me its a way to get my mind off of what happens to be
stressing me out that day.
I also go a lot when I am depressed. It makes me feel good about myself like Im doing something for
myself. I go shopping when I want to treat myself to something special. When I feel I have put so
much time into working that I need a reward.
I want to shop. Its like giving yourself a pat on the back and saying it was worth all the hard work to be
able to shop.
Idea shopping Sometimes I go to look just to get an idea. If I feel like or see a style in a magazine I might go to a
store and try it out just to see how it looks. Its something new and something different and I just want to
try it.
I like to shop because I like to see what the latest fashions are. Sometimes its so hard to keep up
with the latest fashions are. I try my best and hope I stay as hip as possible.
I want to see the new things that come out on the market. Its a way of staying in style. It also tells me
if the clothes I have are out of date or not.
Yes, I thrive on being up on the latest fashions. I always have to make a fashion statement so I am
continuously shopping to buy the new trends.
I like new gadgets, new technology and see the new toys that are out there. It is kind of a hobby.
Role shopping Shopping for others I find to be much more pleasurable. The pleasure I find is trying to determine the
needs of the person Im buying for and then determining a gift that tailors to that persons needs.
I love giving gifts, especially if it something different and unique. Its not just something that you can
get at any other store. I think people know when you spend extra time getting them something that you
know that they really want. This is what I love to do, is give something to someone that you know they
wouldnt purchase for themselves.
I shop more for others than for myself. I especially shop more for my grandchildren and my three
sons. Having my oldest son married and on his own with two daughters, I love to help in anyway
possible.
Shopping for others makes me feel good. It makes me feel good to see how I can help make them
happy. To give someone happiness or to give them something they might not usually get for themselves
make me feel really good.
I love to buy gifts for other people. It makes me feel good to buy something for someone that I know
they are going to like. It is a satisfaction to me. I like to go shopping when it is for other people. I love
that.
Value
I like to turn it into a game, how cheap can I get it. My favorite thing to do is to find something I really
shopping like in the mall, like wet seal or something, that costs $50, and then see how cheap I can find it
somewhere else. I almost always find it for $15 or so. What a feeling when that happens. I live for
those days.
I look at sales and when I go shopping for fun I look at all the sale racks. I like to see whats on sale and
I dig through things.
And it is just like pushing or stretching my money to the boundaries or right to the edge is exciting. It is
exciting to see how much I can buy with what I have.
How is this experience exciting? Its exciting, because you feel like your winning. You feel like I have
control over my money, because I can buy 5 items for $20, rather than 2 items for $20. Thats like the
competitive part of shopping.
Its really fun when you come back with a bunch of goodies for a lot less money than you planned on
spending. Two for ones.
Appendix B. Scale/item measurement properties

Construct Itemsa Coefficient Composite EFA Item Corrected CFA item Squared Scale/item
reliabilityb loading item-total loading multiple mean
correlation correlation
S1c S2d S1 S2 S1 S1 S2 S1 S2 S1 S2 S1 S2
Adventure .86 .86 .88 .92 9.09 9.12
shopping To me, shopping is an adventure .86 .79 .74 .89 .89 .79 .79 3.12 3.17
I find shopping stimulating .67 .74 .81 .88 .94 .78 .89 3.14 3.27
Shopping makes me feel like I am .57 .67 .68 .77 .82 .59 .67 2.83 2.69
in my own universe
Gratification .79 .77 .83 .80 11.92 11.87 M
shopping When Im in a down mood, I go .69 .69 .72 .87 .84 .76 .71 3.56 3.50 .J.
shopping to make me feel better Ar
To me, shopping is a way to relieve .67 .66 .60 .83 .82 .68 .67 3.78 3.75 no
stress ld,
I go shopping when I want to treat .50 .55 .50 .66 .61 .44 .37 4.58 4.63 K.
myself to something special E.
R
Role shopping .83 .84 .86 .88 14.79 14.63 ey
I like shopping for others because .83 .72 .73 .83 .90 .69 .81 5.09 4.96 no
when they feel good I feel good ld
I enjoy shopping for my friends and .80 .74 .75 .91 .89 .83 .80 5.04 5.06 s/
family Jo
I enjoy shopping around to find the .57 .62 .64 .71 .73 .50 .53 4.66 4.61 ur
perfect gift for someone na
Value shopping .85 .87 .88 .90 14.32 14.39 l
For the most part, I go shopping .86 .68 .71 .78 .80 .61 .65 4.58 4.61 of
when there are sales R
I enjoy looking for discounts when I .83 .77 .80 .93 .93 .87 .86 4.99 5.06 et
shop ai
I enjoy hunting for bargains when I .74 .71 .73 .81 .87 .66 .76 4.74 4.72 li
shop ng
79
Social shopping .85 .83 .89 .88 10.47 10.30 (2
I go shopping with my friends or .85 .76 .74 .89 .90 .80 .81 3.47 3.34 00
family to socialize
I enjoy socializing with others when .76 .68 .63 .75 .74 .56 .54 3.76 3.71
I shop
Shopping with others is a bonding .71 .75 .71 .91 .89 .83 .80 3.23 3.26
experience
Idea shopping .84 .87 .88 .90 10.87 10.86
I go shopping to keep up with the .87 .74 .83 .93 .95 .86 .91 3.36 3.37
trends
I go shopping to keep up with the .84 .74 .81 .86 .93 .74 .87 3.65 3.52
new fashions
I go shopping to see what new .70 .62 .65 .72 .69 .52 .47 3.86 3.97
products are available
a
Measurement based on a 7-point Likert scale where 1 = Strongly Disagree and 7 = Strongly Agree.
b 93
See Fornell and Larcker (1981) for details on the composite reliability calculation.
c
S1 = sample 1, calibration sample (n = 266).
d
S2 = sample 2, validation sample (n = 251).
94 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795

References Chin, Wynne, & Todd, Peter A. (1995). On the use, usefulness and ease of
use of structural equation modeling in MIS research: A note of
caution. MIS Quarterly, (June) 237246.
Anderson, James C., & Gerbing, David W. (1982). Some methods for
Churchill, Gilbert A., Jr. (1979). A paradigm for developing better
respecifying measurement models to obtain unidimensional construct
measures of marketing constructs. Journal of Marketing Research,
measurement. Journal of Marketing Research, 19(4), 453461.
16(February), 6473.
Anderson, James C., & Gerbing, David W. (1988). Structural equation
Cope, Nigel. (1996). Retail in the digital age. London: Bowerdean.
modeling in practice: A review and recommended two-step approach.
Cronbach, Lee J., & Meehl, Paul E. (1955). Construct validity in psy-
Psychological Bulletin, 103(3), 411423.
chological tests. Psychological Bulletin, 52(July), 281302.
Babin, Barry J., Darden, William R., & Griffin, Mitch. (1994). Work
Csikszentmihalyi, Mihaly. (1990). Flow: The psychology of optimal ex-
and/or fun: Measuring hedonic and utilitarian shopping value. Journal
perience. New York, NY: Harper & Row.
of Consumer Research, 20(March), 644656.
Dardin, William R., & Ashton, Dub. (1974). Psychographic profiles of
Bagozzi, Richard P. (1980). Causal models in marketing. New York: Wiley.
patronage preference groups. Journal of Retailing, 50(Winter), 99102.
Batra, Rajeev, & Ahtola, Olli. (1991). Measuring the hedonic and utilitar-
Dardin, William R., & Reynolds, Fred D. (1971). Shopping orientations
ian sources of consumer attitudes. Marketing Letters, 2(April), 159
and product usage roles. Journal of Marketing Research, 8(November),
170.
505508.
Bearden, William O., Netemeyer, Richard G., & Teel, Jesse E. (1989).
Dawson, Scott, Bloch, Peter H., & Ridgway, Nancy M. (1990). Shopping
Measurement of consumer susceptibility to interpersonal influence.
motives, emotional states and retail outcomes. Journal of Retailing,
Journal of Consumer Research, 19(March), 473481.
66(4), 408427.
Beatty, Sharon E., & Ferrell, M. Elizabeth. (1998). Impulse buying:
Donovan, Robert J., & Rossiter, John R. (1982). Store atmosphere: An
Modeling its precursors. Journal of Retailing, 74(Summer), 169192.
environmental psychology approach. Journal of Retailing, 58(Spring),
Belk, Russell W. (1984). Three scales to measure constructs related to ma-
3457.
terialism: Reliability, validity, and relationships to measures of happi-
Festinger, L. (1954). A theory of social comparison processes. Human
ness. In Thomas Kinnear (Ed.), Advances in consumer research (Vol.
Relations, 7(1), 117140.
11, pp. 291297). Provo, UT: Association for Consumer Research.
Fischer, Eileen, & Arnold, Stephen J. (1990). More than a labor of love:
Belk, Russell W. (1985). Materialism: Trait aspects of living in the Gender roles and Christmas shopping. Journal of Consumer Research,
material world. Journal of Consumer Research, 12(December), 265 17(December), 333345.
280. Fornell, Claes, & Larcker, David F. (1981). Evaluating structural equation
Bellenger, Danny N., & Korgaonkar, Pradeep K. (1980). Profiling the models with unobservable variables and measurement error. Journal
recreational shopper. Journal of Retailing, 56(Fall), 7792. of Marketing Research, 18(February), 3950.
Bellenger, Danny N., Robertson, Dan H., & Greenberg, Barnett. (1977). Fournier, Susan. (1996). Land Rover North America, Harvard Business
Shopping center patronage motives. Journal of Retailing, School Case 9-596-036. Boston, MA: Harvard Business School Pub-
53(Summer), 2938. lishing.
Bentler, P. M., & Bonnet, Douglas G. (1980). Significance tests and Freud, Sigmund. (1933). New introductory lectures on psycho-analysis.
goodness of fit in the analysis of covariance structures. Psychological New York: W.W. Norton.
Bulletin, 88(3), 588606. Gerbing, David W., & Anderson, James C. (1988). An updated paradigm
Berlyne, Daniel. (1969). Laughter, humor and play. In G. Lindzey & for scale development incorporating unidimensionality and its assess-
E. Aronson (Eds.), Handbook of social psychology (pp. 291297). ment. Journal of Marketing Research, 25(May), 186192.
Reading, MA: Addison-Wesley. Goffman, E. (1959). The presentation of self in everyday life. Garden
Bitner, Mary Jo, Booms, Bernard H., & Tetrault, Mary Stanfield. (1990). City, NY: Doubleday and Company.
The service encounter: Diagnosing favorable and unfavorable inci- Gwinner, Kevin P., Gremler, Dwayne D., & Bitner, Mary Jo. (1998).
dents. Journal of Marketing, 54(January), 7184. Relational benefits in service industries: The customers perspective.
Bloch, Peter H., & Richins, Marsha L. (1983). Shopping without pur- Journal of the Academy of Marketing Science, 26(Spring), 101114. Hair,
chase: An investigation of consumer browsing behavior. Advances in Joseph F. Jr., Anderson, Rolph E., Tatham, Ronald L., & Black, William
Consumer Research, 10, 389393. C. (1998). Multivariate data analysis. Upper Saddle River,
Bloch, Peter H., Ridgway, Nancy M., & Dawson, Scott A. (1994). The NJ: Prentice Hall.
shopping mall as consumer habitat. Journal of Retailing, 70(1), 23 Hawes, Jon, & Lumpkin, James R. (1984). Understanding the outshopper.
42. Journal of the Academy of Marketing Science, 12(4), 200218.
Bloch, Peter H., Ridgway, Nancy M., & Sherrell, Daniel L. (1989). Hirschman, Elizabeth C. (1980). Innovativeness, novelty seeking and
Extending the concept of shopping: An investigation of browsing consumer creativity. Journal of Consumer Research, 7(3), 283291.
activity. Journal of the Academy of Marketing Science, 17(1), 13 Hirschman, Elizabeth C., & Holbrook, Morris B. (1982). Hedonic con-
21. sumption: Emerging concepts. Journal of Marketing, 46(Summer),
Bloch, Peter H., Sherrell, Daniel L., & Ridgway, Nancy M. (1986). Con- 92101.
sumer search: An extended framework. Journal of Consumer Hoffman, Donna L., & Novak, Thomas P. (1996). Marketing in
Research, 13(June), 119126. hypermedia computer-mediated environments: conceptual
Bollen, Kenneth A. (1989). Structural equations with latent variables. foundations. Journal of Marketing, 60(July), 5068.
New York: Wiley. Holbrook, Morris B., & Hirschman, Elizabeth C. (1982). The experiential
Brady, Diane. (2000). Why service stinks. Business Week, 3704(October aspects of consumption: Consumer fantasies, feelings, and fun.
23), 118128. Journal of Consumer Research, 9(March), 132140.
Bruning, James L., & Kintz, B. L. (1977). Computational handbook of Hudson, Laurel A., & Ozanne, Julie L. (1988). Alternative ways of
statistics. Glenview, IL: Scott Foresman and Company. seeking knowledge in consumer research. Journal of Consumer
Burke, Raymond. (1997). Do you see what I see? The future of virtual Research, 14(March), 508522.
shopping. Journal of the Academy of Marketing Science, 25(4), 352 Huizinga, Johan. (1970). Homo Ludens: A study of the play element in
361. culture. New York: Harper and Row.
Buss, Dale D. (1997). Entertailing. Nations Business, 85(12), 1218. Jarboe, Glen R., & McDaniel, Caral D. (1987). A profile of browsers
Byrne, Barbara. (1998). Structural equation modeling: Basic concepts, in regional shopping malls. Journal of the Academy of Marketing
applications, and programming. Mahwah, NJ: Lawrence Erlbaum. Science, 15(1), 4653.
 M.J. Arnold, K.E. Reynolds / Journal of Retailing 79 (2003) 7795 95
Oliver, Richard L., (1997). Satisfaction: A behavioral perspective on the
consumer. Boston, MA: Irwin, McGraw-Hill.
Jones, Michael, Reynolds, Kristy E., Weun, Seungood, & Beatty, Sharon
Oliver, Richard L., & Bearden, William O. (1985). Crossover effects in
E. (in press). The product-specific nature of impulse buying tendency.
the theory of reasoned action: A moderating influence attempt.
Journal of Business Research.
Journal of Consumer Research, 12(December), 324340.
Jreskog, K., & Srbom, D. (1993). LISREL 8: Users reference guide.
Otnes, Cele, & McGrath, Mary Ann. (2001). Perceptions and realities of
Hillsdale, NJ: Lawrence Erlbaum. male shopping behavior. Journal of Retailing, 77(Spring), 111 137.
Kowinsky, William S. (1985). The malling of America. New York: William Peter, J. Paul. (1981). Construct validity: A review of basic issues and
Morrow and Company, Inc. marketing practices. Journal of Marketing Research, 18(May), 133
Langrehr, Frederick W. (1991). Retail shopping mall semiotics and 145.
hedonic consumption. Advances in Consumer Research, 18, 428 Peterson, Robert A. (2000). A meta-analysis of variance accounted for
433. and factor loadings in exploratory factor analysis. Marketing Letters,
Lee, Euehun, Moschis, George P., & Mathur, Anil. (2001). A study of life 11(3), 261275.
events and changes in patronage preferences. Journal of Business Punj, Girish, & Staelin, Richard. (1983). A model of consumer
Research, 54(October), 2538. information search for new automobiles. Journal of Consumer
Lincoln, Yvonne, & Guba, Egon. (1985). Naturalistic inquiry. Beverly Research, 9(March), 366380.
Hills: Sage. Punj, Girish, & Stewart, David. (1983). Cluster analysis in marketing
MacCallum, Robert. (1986). Specification searches in covariance structure research: review and suggestions for application. Journal of
modeling. Psychological Bulletin, 100(1), 107120. Marketing Research, 20(May), 134148.
MacCallum, Robert, Roznowski, Mary, & Necowitz, Lawrence B. (1992). Reynolds, Kristy E., & Beatty, Sharon E. (1999). A relationship customer
Model modifications in covariance structure analysis: The problem of typology. Journal of Retailing, 75(4), 509523.
capitalization on chance. Psychological Bulletin, 111(3), 490504. Rook, Dennis W. (1987). The buying impulse. Journal of Consumer
Mathwick, Charla, Malhotra, Naresh, & Rigdon, Edward. (2001). Ex- Research, 14 (September). 189199.
periential value: Conceptualization, measurement, and application in Roy, Abhik. (1994). Correlates of mall visit frequency. Journal of Retail-
the catalog and Internet shopping environment. Journal of Retailing, ing, 70(Summer), 139161.
77(Spring), 3956. Sherry, John F, Jr. (1990). A Sociocultural analysis of a midwestern
McClelland, David C. (1961). The achieving society. Princeton, NJ: Van American flea market. Journal of Consumer Research, 17(June), 13
Nostrand. 30.
McGuire, William. (1974). Psychological motives and communication Sorokin, P. A. (1950). Altruistic love: A study of American good neighbors
Gratification. In J. F. Blumer & Katz (Eds.), The uses of mass commu- and Christian saints. Boston: Beacon Press.
nication: Current perspectives on gratification research (pp. 106 Steenkamp, Jan-Benedict E. M., & Baumgartner, Hans. (1998). Assessing
167). Beverly Hills: Sage. measurement invariance in cross-national consumer research. Journal
Midgley, David F. (1977). Innovation and new product marketing. New of Consumer Research, 25(1), 7891.
York: Halsted Press, Wiley. Stephenson, P. Ronald, & Willett, Ronald P. (1969). Analysis of
Midgley, David F., & Dowling, Grahame R. (1978). Innovativeness: consumers retail patronage strategies. In P. R. McDonald (Ed.), Mar-
The concept and its measurement. Journal of Consumer Research, keting involvement in society and the economy (pp. 316322).
4(March), 229242. Chicago: American Marketing Association.
Miller, Daniel. (1998). A theory of shopping. Ithaca, NY: Cornell Uni- Stone, Gregory P. (1954). City shoppers and urban identification: Ob-
versity Press. servation on the social psychology of city life. American Journal of
Milligan, Glenn. (1980). An examination of the effect of six types of error Sociology, 60(July), 3645.
perturbation of fifteen clustering algorithms. Multivariate Behavioral Tauber, Edward M. (1972). Why do people shop? Journal of Marketing,
Research, 15, 225238. 36(October), 4649.
Milligan, Glenn, & Cooper, Martha C. (1985). An examination of proce- Tian, Kelly Tepper, Bearden, William O., & Hunter, Gary L. (2001).
dures for determining the number of clusters in a data set. Psychome- Consumers need for uniqueness: Scale development and validation.
trika, 50(2), 159179. Journal of Consumer Research, 28(June), 5066.
Moschis, George P. (1976). Shopping orientations and consumer use of Thompson, Craig J., Locander, William B., & Pollio, Howard R. (1990).
information. Journal of Retailing, 52(Summer), 6170. The lived meaning of free choice: An existential-phenomenological
Neisser, Marianne. (1973). The sense of self through giving and receiving. description of everyday consumer experiences of contemporary mar-
Social Casework, 54(5), 294301. ried women. Journal of Consumer Research, 17(December), 346
Netemeyer, Ricard G., Durvasula, Srinivas, & Lichtenstein, Donald R. 361.
(1991). A cross-national assessment of the reliability and validity of Wakefield, Kirk L., & Baker, Julie. (1998). Excitement at the mall:
the CETSCALE. Journal of Marketing Research, 28(August), 320 Determinants and effects on shopping response. Journal of Retailing,
327. 74(Fall), 515540.
Netemeyer, Ricard G., Johnston, Richard G., & Burton, Mark W. (1990). Wallendorf, Melanie, & Arnould, Eric. (1991). We gather together: Con-
Analysis of role conflict and role ambiguity in structural equations. sumption rituals of thanksgiving day. Journal of Consumer Research,
Journal of Applied Psychology, 75(2), 148158. 18(June), 1331.
Novak, Thomas P., Hoffman, Donna L., & Yung, Yiu-Fai. (2000). Mea- Westbrook, Robert A., & Black, William. (1985). A motivation-based
suring the customer experience in on-line environments: A structural shopper typology. Journal of Retailing, 61(Spring), 78103.
modeling approach. Marketing Science, 19(Winter), 2242. Zaichowsky, Judith Lynne. (1985). Measuring the involvement construct.
Nunnally, Jum C., & Bernstein, Ira H. (1994). Psychometric theory (3rd Journal of Consumer Research, 12(December), 341352.
ed.). New York: McGraw-Hill.
OGuinn, Thomas C., & Faber, Ronald J. (1989). Compulsive buying: A
phenomenological exploration. Journal of Consumer Research, 16(2),
147158.