h 59 (2006) 535–548

Journal of Business Researc

Consumer decision-making styles and mall shopping behavior: Buildingtheory using exploratory data analysis and the comparative method

Scarlett Wesley a,⁎, Melody LeHew b,1, Arch G. Woodside c,2

a University of Kentucky, College of Agriculture, Department of Merchandising, Apparel, and Textiles, 318 Erikson Hall, Lexington, KY 40506, United Statesb Kansas State University, College of Human Ecology, Department of Apparel, Textiles, and Interior Design, 217 Justin Hall, Manhattan, KS 66502, United States

c Boston College, Carroll School, Department of Marketing, 140 Commonwealth Avenue, Chestnut Hill, MA 02467, United States

Received 1 October 2005; received in revised form 1 December 2005; accepted 1 January 2006

Abstract

This article assesses how consumers' decision-making styles relate to their shopping mall behavior and their global evaluations of shoppingmalls. Based on exploratory data analysis including the use of the comparative method, the article provides a theoretical model of antecedentsand consequences of consumer-decision making styles. Data for this report come from personal face-to-face mall intercepts of shoppers(n=527) in two super-regional (West Edmonton Mall and Mall of America in Bloomington) and two regional (Pier 39 in San Francisco andForum Shops in Las Vegas) malls. The EDA results support a complex view of the antecedents and consequences of consumer decision-making styles. The article concludes with specific suggestions for extending psychological theory of shopping behavior and advancing strategicmall-retailing strategies.© 2006 Elsevier Inc. All rights reserved.

Keywords: Consumer decision-making styles; Exploratory data analysis; Mall shopping activities; Satisfaction; Intention-to-return

1. Introduction: toward a theory of CDM styles

Similar to the work on cognitive styles in psychology (e.g.,see Sternberg and Grigorenko, 1997) and beginning with theseminal work of Stone (1954), the literature consumerdecision-making (CDM) styles have a fairly long but variedhistory in the context of theory and research in retailing.Sproles and Kendall (1986) define a CDM style as “a mentalorientation characterizing a consumer's approach to makingchoices.” The work of Westbrook and Black (1985) and Hiu,Siu, Wang, and Chang (2001) include CDM styles literaturereviews from 1954 to 1985 and 1986 to 2001, respectively.The substantial majority of studies in both these early andlater research streams focus on confirming the existence of

⁎ Corresponding author. Tel.: +1 859 257 7778; fax: +1 859 251 4917.E-mail addresses: scarlett.wesley@uky.edu (S. Wesley),

lehew@humec.ksu.edu (M. LeHew), woodsiar@bc.edu (A.G. Woodside).1 Tel./fax: +1 785 532 1321/3796.2 Tel./fax: +1 617 552 3069/6677.

0148-2963/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.jbusres.2006.01.005

styles of thinking applied to consumer shopping cognitionsand that segmenting consumers (e.g., clustered) into a limitednumber of meaningful groups by their responses to CDMstyle inventories is possible and useful. Few studies attemptto thoroughly explore the antecedents and consequences ofCDM styles (Hiu et al., 2001 is an exception to thisobservation).

This article proposes and illustrates the use of an exploratorydata analysis (EDA) approach for developing a theory ofantecedents and consequences of CDM styles. The researchincludes examining the propositions that (1) shopping contexts(e.g., shopping in regional versus local area malls) theoreticallyshould not affect the structure of CDM styles; (2) at least someCDM styles moderately relate with each other; (3) mostshoppers identify themselves as applying more than one butless than most CDM styles (i.e., shoppers recognize that somespecific styles do not apply to their shopping orientations); (4)demographic variables affect CDM styles; (5) CDM stylesrelate with shoppers' planned expenditure levels; (6) CDMstyles relate directly to global satisfaction with mall shopping;

Table 1Characteristics of eight consumer decision-making styles

Decision style

1. Perfectionist/high quality-conscious consumer: decision style of consumerswho systematically search for the best quality products possible.

2. Brand consciousness/price equals quality: decision style of consumersconcerned with getting the most expensive, well-known brands.

3. Novelty and fashion conscious: decision style of consumers who like newand innovative products and who gain excitement from seeking out newthings.

4. Recreational and shopping conscious: decision style of consumers who takepleasure in shopping and who shop just for the fun of it.

5. Price conscious/value for the money: decision style of consumers who areconcerned with getting lower prices.

6. Impulsiveness/careless: decision style of consumers who tend to buyspontaneously and who are unconcerned about how much money theyspend.

7. Confused by overchoice: decision style of consumers who feel they have toomany brands and stores to choose from and who likely experienceinformation overload in the market.

8. Habitual/brand loyal: decision style of consumers who shop at the samestores and tend to buy the same brands each time.

Source: Adapted from Sproles and Kendall (1986).

536 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

and (7) CDM styles affect the activities shoppers engage in.Findings from the EDA of survey responses from samples ofshoppers in four North American malls support the first sixbut not the sixth proposition. The results of the study supportsa complex view that consumers are capable and prone toapplying more than one CDM style when shopping and thatthe impact of CDM styles on shopping activities is moreindirect than direct.

The findings from the present study confirm and extendSproles and Kendall's (1986, p. 268) previously unexploredspeculation that a person does not follow one CDM style in allshopping decisions. The most useful view regarding CDMstyles may be that many consumers adopt two to threeapproaches to making choices and rarely apply all stylesproposed theoretically and found in shopping research; thepresent article confirms the possibility that a substantial share ofconsumers are not oriented strongly to any one CDM style.Thus, a substantial share of shoppers exists that rarely use eachof the eight to ten CDM styles (e.g., Sproles and Kendall, 1986;Tai, 2005).

This article includes the following sections. Following thisintroduction, Section two proposes a theory of CDM stylesincluding antecedents and consequences relating to styles.Section three briefly summarizes the empirical method andapproach to analyzing the data for probing the theory. Theresearch method includes a survey data collection instrumentand the use of a mall-intercept face-to-face interview method;the analyses include exploratory data analysis (EDA) includinga comparative method approach to survey data. Section fourpresents findings. Section five discusses conclusions andimplications for further theory development and retail manage-ment practice. Section six closes the article with limitations, andsuggestions for further research.

2. Theory of CDM styles

Retailers and shopping mall developers often seek to learnhow and why people shop. This search results in severalparadigm proposals of different shopping typologies (e.g.,Bellenger and Korgaonker, 1980; Darden and Reynolds, 1971;McDonald, 1993; Sharma and Levy, 1995; Stone, 1954; Walshet al., 2001a,b; Wang et al., 2004; Westbrook and Black, 1985).These studies are successful in increasing knowledge of theshopping orientations of consumers. Sproles (1985) and Sprolesand Kendall (1986) provide the Consumer Styles Inventory(CSI), which is an early attempt to systematically measureshopping orientations using decision-making orientations.Sproles (1985) provides eight central decision-making dimen-sions to explain why shoppers behave in certain ways. One ofthe most important assumptions of this approach is that eachindividual consumer has a specific decision-making styleresulting from a combination of their individual decision-making dimensions.

The theoretical assumption behind Sproles and Kendall's(1986) ideas about CDM styles is that consumers have eightdifferent decision-making dimensions that determine theshopping decisions they make. Sproles and Kendall (1986)

identify and validate eight different dimensions from a sampleof 482 U.S. high school students. The participants were askedclosed-end questions to assess their decision-making style forpersonal products (i.e., clothing, cosmetics and hairdryers). Intotal 40 items pertaining to affective and cognitive orientationsin decision-making was the basis from which eight potentialstyles affecting behavior were identified (see Table 1).

Fig. 1 visualizes theoretical proposals of antecedents andconsequences of CDM styles. For brevity Fig. 1 alsosummarizes the test results of hypotheses that this articledescribes in the Findings Section below. Fig. 1 includes thefollowing hypotheses relating CDM styles and mall shoppingbehavior.

H1. Type of mall (super or mega mall versus smaller regional orneighborhood mall) does not affect CDM styles—the predictionis that findings will support a null relationship. Rationale: CDMstyles are thinking styles that are preferred ways of using theabilities that an individual develops over several years, that is,CDM styles serve to bridge personality and consumers' use ofrules or heuristics stored in memory (cf. Bettman and Zins,1977; Sternberg and Grigorenko, 1997). Consequently, theexpectation is that type of shopping mall does not affect thestructure of CDM styles.

H2. Demographics affect CDM styles. Rationale: findings fromprior research support the view that gender, age, and incomeinfluence the adoption of specific CDM styles (e.g., seeKamaruddin and Mokhlis, 2003; Walsh et al., 2001a,b). Forexample, consumers 18- to 24-years-old are “more likely thanother consumers to buy a product on the spur of the moment andchange brands if the mood strikes” (Weiss, 2003, p. 31),whereas consumers 27- to 39-years old are “looking forproducts that seem less mass-marketed and more retro, whilealso being affordable” (Wiggins, 2004, p. 37).

Consumer Decision-Making Styles

PerfectionistBrand consciousnessNovelty & fashion consciousRecreational & shoppingPrice consciousImpulsiveness/carelessConfused by over choiceHabitual / brand loyal

Mall Type

�

Super regionalRegional/Neighborhood

Demographic / SESGender

IncomeEducationAge

Prime Reason Visitingin AreaResidentShoppingVacationBusinessFamily

Prime ReasonFor Mall Visit

EntertainmentEatingSocializing

Activities DuringVisit in MallShoppingBrowsingMoviesEatingSee exhibitsWalking SocializeAttractions / rides

Satisfaction / Dissatisfaction

My experience Merchandise

Mall environment

Store varietyGlobal

Intention tovisit mall

Come back to cityspecifically to

visit mall

2

3

4

67

8ns

9

10 phi = .42, p < .01

PlannedExpenditure

Level 11

5 phi = .32, p < .01

1 (ns)

Dictionary: ns = not significant

12r = .18p < .01

13

2 = .04,p < .000

ns

15phi = .44p < .000

14 (see Fig. 5)

(see Fig. 3)

phi = .43, p < .000 phi = .33p < .000

ns

ns

Learning StylesSix styles (see Kendall,

Sproles, and Sproles, 1990)

16(KSS support)

17 Untested

18

15

η

η2 = eta square

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

Shopping

�

�

�

�

Prices

Fig. 1. Theory of antecedents and consequences of consumer decision-making styles.

537S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

H3. Prime reason for visiting in the area/community in whichthe mall is located where the interview takes place with aninformant does not relate to CDM styles. The rationale is thesame as the one for H1. Because styles are enduring over yearsor even decades, a particular reason for an informant being inthe geographic area near where the mall is located should havelittle bearing on the informants CDM style.

H4. CDM styles affect shoppers' activities (e.g., shopping tobuy, browsing, eating, socializing, going to the movies if atheater is available) during mall visits. Rationale: shoppingactivities represent one category of choices implemented byshoppers. CDM styles are a subset of the general construct ofstyle which Guralnik (1976, p. 1415) defines as a distinctive orcharacteristic manner or method of acting or performing.Shopping activities done in malls is one form of performing (cf.Deighton, 1992).

H5. How much a shopper plans to spend in a mall affectswhich activities she/he performs in the mall (e.g., plans forspend more than $100 relates more to shopping thanbrowsing or socializing in the mall). While H5 may appearobvious with no need to substantiate, empirical support for H5

helps to support nomological validity (Peter, 1981) for thetheory-data analysis is necessary to confirm that formalhypotheses that are obviously supportable empirically areindeed supportable external to their assertion by a researcher.Empirical measurement of relationship strength of obviouslysupportable hypotheses provides some indication of an upper

bound of effect size of relationships claimed to be obviouslysupportable.

H6. Some specific CDM styles relate more than others tospecific primary reasons for the shopper visiting a mall. Forexample, the following specific statement reflects H6: a greatershare of shoppers highly-oriented to being recreational andshopping conscious visit a shopping mall primarily forshopping compared to the share of shoppers highly-orientedto different CDM styles. Rationale: as personality-relatedvariables CDM styles antecede primary reasons for a givenmall visit; in a general way-across a sample of informants—such fundamental ways of approaching contexts shoppers findthemselves in should affect these shoppers' conscious thinkingabout why they are in such contexts. “I must of have come to themall to see the latest fashions,” is how a shopper highly noveltyand fashion conscious may come to conclude that her or hisprimary reason for being in the mall was for shopping (cf.Martin, 2005).

H7. The prime reason for a shopper's mall visit influences theactivities the visitor reports engaging in while visiting the mall.Rationale: going to a shopping mall versus staying home orgoing to other destinations requires some amount of consciousthought that a shopper is likely to be able to retrieve and reportin response to aided recall questioning. However, Weick's view(1995) that people make sense of their actions retrospectivelysupports the opposite flow of influence-activities done influ-ences primarily reason. Using survey data collected while

538 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

shoppers are in malls, the conservative bi-directional view thatprime reason relates to activities done may be more appropriatethan a one-directional hypothesis. Thus, while H7 is one-directional, the data analysis includes association only statisticaltests.

H8. Engaging in specific activities or combinations of activitiesin a shopping mall results in higher satisfaction by the shopperwith her or his mall visit than engaging in other specificactivities or combinations of activities. Rationale: mall visitorsshopping versus browsing only are likely to be more satisfiedwith their visit because the result of the visit includes a valuedtangible outcome; shoppers engaging in several activities (e.g.,eating, socializing, shopping, and visiting attractions) are likelyto conclude that their visits were highly satisfying versusvisitors engaging in just one of these activities—for a givenmall visit, doing lots of different activities generates highersatisfaction than performing only one activity.

H9. Mall visit satisfaction influences intention to visit the mall.Rationale: a substantial stream of theory and research supportsthe view that satisfaction with a service or product affectsintention to buy and use the service or product again (e.g., seeHoward and Sheth, 1969; Oliver et al., 1997).

H10. Mall type affects shoppers' activities while visiting themall. Rationale: most shoppers only report doing activities thata mall offers. For example, some mall visitors will report inengaging in water sports and amusement rides in super regionalmalls offering such activities but few, if any, will provide similarreports in smaller malls not offering such activities. Empiricallyexamining this intuitively obvious hypothesis offers a rule-of-thumb of the upper limit of effect sizes that may be expectedwhen examining associations among operational constructs insurvey research.

H11. Consumer decision making styles affect shoppers'planned expenditure levels. Rationale: CDM styles, such asperfectionist and brand consciousness, relate to seeking highquality and equate to paying high prices with high quality; otherCDM styles, such as price conscious and value for money, relateto an unwillingness to pay high prices. The expectation is thatshoppers highly oriented to acquiring high priced products andservices are more likely to plan on higher expenditures thanshoppers highly oriented to be unwilling to pay high prices.

H12. Planned expenditure level positively affects satisfaction.Rationale: assuming that planning to spend leads to expendi-tures, spending money to acquire products and services is fun—a rallying cry for many consumers. H12 is likely to be highlysupported among high versus low income mall shoppers—planning to spend is especially fun if the mall visitor has highdiscretionary funds available for spending.

H13. Mall type affects planned expenditure level. Rationale:some malls may focus on activities that relate more closely withhigh expenditures compared to other malls. For example, smallregional malls not providing amusement rides, movies, and fewrestaurants may relate particularly to shopping versus superregional malls that encourage varied activities including

walking and socializing; consequently, average planned expen-ditures may be higher for small versus large malls.

H14. Demographics affect planned expenditure level. Ratio-nale: high versus low income mall visitors are more likely toplan high versus low expenditures because they have greaterdiscretionary funds for shopping. Females versus males visitmalls on shopping trips versus alternative reasons and shoppingrelates positively to planned expenditures. The main effect foreducation on planned expenditure level is likely to be negative(controlling for income) because the enjoyment resulting frommall shopping and expenditures is greater for lower income mallvisitors; mall visitors with higher education are more likely toplan high expenditures elsewhere (e.g., foreign travel, invest-ments, and health club memberships).

H15. Demographics moderate the influence of consumerdecision-making styles on planned expenditures. Rationale:high income enables consumers highly perfectionist and brandconscious to plan high expenditures; female, high income, highperfectionist are more likely to plan high expenditures versusmale, high income, high perfectionist—and visitors with bothof these profiles have higher planned expenditures than othermall visitors. While intuitively appealing, this hypothesis came-to-mind only after examining the data and after failing to findsupport for H4. As Tukey (1977) likens EDA to detective work,the overall research aim for the present study includesinvestigating all reasonable direct and indirect influences ofCDM styles on mall shopping behavior, as well as the direct andindirect influences of latent variables influencing CDM styles.

H16 and H17. Learning styles affect CDM styles anddemographics affect learning styles. Fig. 1 includes these twohypotheses for completeness and to reflect Sproles and Sproles(1990) finding of associations among 21 of 48 possible pairs ofrelationships among six learning styles and eight CDM styles.Learning styles include six characteristics: serious, analyticallearner; active practical learner; observation-centered learner;passive, accepting learner; concrete, detailed, fact-orientedlearner; nonadaptive, struggling learner (Sproles et al., 1987;Sproles and Sproles, 1990). Possibly because Sproles andSproles (1990) survey administration includes only secondaryschool students, their study includes no examination ofdemographic influences on learning styles. However, futureresearch is likely to confirm that demographics relate to learningstyles. The present article does not include empirical examina-tions of H16 and H17.

H18. CDM styles directly affect global satisfaction withshopping in a mall, that is, consumers exhibiting very highlevels for some CDM styles (e.g., perfectionist/high quality-conscious consumers have higher levels of satisfaction withmall shopping than consumers exhibiting very low levels).Rationale: very high versus very low levels of some CDM styleslikely reflect a general passion for shopping—shopping mallvisits manifest such passion. Only consumers with very highlevels of certain CDM styles are likely to respond to the call-to-arms of “shop till you drop;” these are the shoppers whoperceive inherent joy of visiting shopping malls. In a field

539S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

relating to retailing, health services research supports a strongassociation with decision-making styles and patient satisfaction(e.g., patients' satisfaction with care relates positively with theirdegree of involvement in decision-making during visits (Joos etal., 1993).

3. Method

The empirical method for this study adopts an exploratorystance with respect to both data collection and analysis for tworeasons: (1) most research reports on CDM styles followSproles and Kendall's (1986) approach in applying their stylesinventory on data from high school and college students and/or(2) do not examine antecedents and consequences associatingwith high versus low levels on unique style factors. Becausehigh school and college students usually come from a verynarrow age range and their CDM styles are likely to be lesscrystallized at this stage than later in life (cf. Sears, 1986), thepresent study focuses on adult consumers ranging in age from18 to 85 plus. Wells' (1993) view is relevant here, “This is notto say that findings based on students are always wrong. It isonly to say that findings based on students are always suspect.”Also, data were collected in (shopping mall) environments thatrelate to shopping-environments likely to be recently andcurrently stimulating conscious thinking that relates to CDMstyles—rather than classroom settings.

3.1. Survey instrument

The survey instrument used for this study was the Sprolesand Kendall (1986) “Consumer Styles Inventory” (CSI). TheCSI includes 39 items that cover the eight styles that Table 1lists. The questionnaire appears in booklet form with theshopping styles being measured using five-point Likert-typescales. On this five-point scale, responses range from stronglyagree (1) to strongly disagree (5). The survey instrumentincludes 25 yes/no questions on activities the informant did orreports will do during the visit to the mall; 14 satisfactionquestions with possible responses for each question rangingfrom satisfied, somewhat satisfied, undecided, somewhatdissatisfied, and dissatisfied. The surveys asks the informantto identify the primary reason for visiting in the city area wherethe shopping mall is located—the list of possible reasonsinclude shopping, vacation, business, family, and other. Thesurvey includes asking the informant, “Today, what is yourprimary reason for visiting [mall name] with five optionalanswers: shopping, eating, entertainment, socializing, other.The survey includes asking informants to provide an approx-imate amount of money she/he plans on spending in the mall onthis visit—using four options ranging from $1–50, $51–100,$100–250, and more than $250. Informants were asked thefollowing global satisfaction question, “From my experience inthis shopping center today, I am…” (satisfied to dissatisfiedusing a 1 to 5 scale). Awillingness to return question to the mallwas used to as a surrogate for measuring intention; the surveyasks, “Would you ever come back to the (city name) area tospecifically visit (mall name). The survey includes demographic

questions on gender, age, ethnic origin, martial status,education, occupation, current work status, and householdincome before taxes.

3.2. Surveying contexts

The study design follows Cook and Campbell's (1979)recommendations for deliberate sampling to achieve heteroge-neity to reflect diversity on presumptively important dimensions(e.g., super regional versus smaller malls; informants living inwidely diverse areas). Thus, the study's design permits probingof the view that the existence of specific CDM styles isindependent of shopping mall contexts and samples fromdiffering populations.

The shopping malls of interest to this study were the WestEdmonton Mall, the Mall of America, The Forum Shops, andPier 39. West Edmonton Mall and Mall of America are superregional malls that are considered to be anchor traveldestinations (i.e., mentioned as a primary destination by asubstantial share of mall visitors). These malls draw visitorsfrom around the world.

Listed in the Guinness Book of World Records as the “largestshopping center in the world”, the West Edmonton Mall covers5.3 million square feet of shopping and entertainment space andis currently listed as Alberta's number one tourist attraction(Fast Facts, 2002). Developers of the mall sought to create amixed shopping and entertainment environment powerfulenough to become a major tourist destination (Getz, 1993).

Located in Bloomington, Minnesota, the Mall of America isthe United State's largest retail and entertainment complex,attracting more visitors annually than Disney World (Mall ofAmerica Continues Its Nine-Year Reign, 2001). Even after tenyears in operation, the Mall of America is a major touristdestination, attracting around 43 million visitors a year andaverages sales around $860 million per year (Lisanti, 1998).

The remaining two malls also attract tourist shoppers, butthese smaller malls may not be the primary attractiondestination for most of their visitors. The Forum Shops andPier 39 are centers that are located nearby very popular traveldestinations: Las Vegas, Nevada and San Francisco, California,respectively. The Forum Shops in Las Vegas is one of theearliest mall-as-theme-park creations, and it draws around$1200 per square foot in annual sales, compared with a nationalaverage of around $300 (Ginsburg, 1999). The Forum Shopshas a unique atmosphere that simulates ancient Roman streets.

The final shopping center of interest was Pier 39 located inSan Francisco, California. Pier 39 is San Francisco's numberone tourist attraction with approximately 10.5 million visitorseach year (Pier 39 Press Priorities). Located just two blocksfrom Fisherman's Wharf, Pier 39 provides stunning views of theGolden Gate Bridge, Alcatraz, Angel Island, and the SanFrancisco Bay (Pier 39 Press Priorities).

3.3. Procedure for data collection

Prior to data collection, the senior management of each mallwas contacted to request participation in the study. A copy of

540 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

the survey and a short proposal were sent to each shoppingcenter. Each senior management was offered a copy of theresults of the study in return for their cooperation. Mallexecutives agreed to allow the researchers to collect data usingthe mall intercept technique. The researchers were positioned ina central location of each shopping center, and an informationalsign detailed the study to visitors. Visitors were asked toparticipate in the study as they passed by the collection location.The questionnaires were self-administered by the respondents,and data were collected at different days of the week anddifferent times of the day to improve randomness. Eachrespondent was given a participation gift of five dollars aftercompleting the questionnaire. The final customer sample was527.

3.4. Steps in the exploratory data analyses

Even with the hypotheses in theory of antecedents andconsequences of CDM styles (see Fig. 1), this study adopts anEDA stance because of the limited prior research of analyzingmultiple samples of adults and absence in prior work on thenomological (lawlike) validity of CDM styles (i.e., investigat-ing both the theoretical relationships among different constructsand the empirical relationships among measures of thosedifferent constructs, see Peter, 1981). The goal of EDA is touncover patterns in the data. “Tukey often likened EDA todetective work. The role of the data analyst is to listen to thedata in as many ways as possible until a plausible ‘story’ of thedata is apparent, even if such a description would not be borneout in subsequent samples” (Behrens, 1997, p. 132).

Graphical analysis is central to EDA (Behrens, 1997; Tufte,2003; Tukey, 1977). Also, EDA recognizes that outlying datatell a different story from the rest of the data, “and to try tosummarize all the data with a single model or statistic leads to acase of combining apples and oranges” (Behrens, 1997, p. 144).Consequently, the data analyses in the present study include theuse of the comparative method (Ragin, 1987) in case studyresearch-examining the stream of antecedents and conse-quences of individual informants among the total samples ofinformants to deepen the understanding of the nomological net(see Cronbach and Meehl, 1955) of relationships that includeCDM styles.

For the study the EDA includes exploratory factor analysis(EFA, see Fabrigar et al., 1999) of the 39 items in the Sproles(1985) eight factor consumer style inventory (CSI). This studyfollows the recommendations of Fabrigar et al. (1999) forconducting EFA. For example, the analysis includes a numberof factor solutions ranging from eight to twelve and the resultingfactor structures were examined to learn if the rotated factorswith the largest explained variance included the same factorstructure (i.e., same items in the same factors) as Sproles (1985)and Sproles and Kendall (1986). Both the scree test and theKaiser criterion was considered in deciding on factor structures(see Fabrigar et al., 1999). Oblique (quartimin rotation) wasapplied for final output of the factor structure becauseorthogonal (e.g., varimax) rotations constrain factors to beuncorrelated and Sproles and Kendall (1986) point out that

some CDM styles are likely to be moderately correlated.(Oblique rotations do not require factors to be correlated; if thesolution with the best simple structure involves orthogonalfactors, a successful oblique rotation provides estimates of thecorrelations among factors close to zero.) Fabrigar et al. (1999)propose that when the average communalities are near. 70 foritems in a common factor, a sample size of 100 might well beadequate; given such results, the analyses includes examiningthe rotated factor structure for the informant sample for eachmall as well as total sample of informants across all four malls.The Findings Section includes an exhibit of statisticallysignificant correlations among the resulting CSI measuresalong with the coefficient alphas for the eight measures.

The EDA includes a tipping point (Gladwell, 2000) analysisto examine both linear and curvilinear relationships of CDMstyles with antecedents and consequences. The tipping pointanalysis for this study applies McClelland's (1998) approach ofgrouping informants in five groups about equal in size fromvery low, low, medium, high, and very high for each measuredCDM style—thus, this tipping point analysis assumes anordinal, rather than cardinal, stance toward the strongly-disagree to strongly-agree five-point value assignments in theinformants style self-reports. The tipping point analysisembraces the view that certain combinations of levels ofdemographics (e.g., middle age with high income) joining witha relatively high (top 20%) CDM style affects shoppingbehavior plans and consequences rather than solely CDMstyles.

4. Findings

Table 2 shows the factor loadings for items loading above.60 for the total sample rather than the less stringentrequirement of .40 often used with EFA. The findings includebreakouts for informant responses for each mall as well as thetotal sample of informants. The first eight factors of a tenfactor solution explains 57% of the variation in the data forthe CSI items, somewhat higher than the 46% level thatSproles and Kendall (1986) report. For the total sample theitems load on the same factors as Sproles (1985) and Sprolesand Kendall (1986) identify to be CDM styles. The itemloadings per factor for each mall closely follow the overalltotal sample except for factor 8.

Fig. 2 includes coefficient alphas and correlations among thefactors (using only the items per factors in Table 1). Thefindings in Fig. 2 support the view that some of the CDM stylesare moderately correlated and that forcing an orthogonalsolution is misleading in examining CDM styles.

H1. Mall type and CDM styles

The findings in Table 2 support the view that CDMstyles are independent of the mall shopping context. For19 of the 28 items the loadings of items occur for thesame factor for at least three of the four malls. Nearlyidentical patterns of significant correlations among itemsoccur for each factor across the four malls (not shown).

Table 2Consumer style characteristics for total sample and by mall

Consumer styles(variance explained)

Total WestEdmonton

Mall ofAmerica

Pier39

Forumshops

Factor 1 — brand conscious, “price equals quality” (15%)The most advertised brands are

usually very good choices71 63 75 72 45

I prefer buying the best sellingbrands

69 69 – 68 –

Nice department and specialtystores offer me the bestproducts

65 56 – 69 –

The higher the price of theproduct, the better the quality

61 64 – 71 –

Factor 2 — perfectionists/high quality conscious (11%)I make a special effort to choose

the very best quality products79 80 81 68 82

In general, I usually try to buythe best overall quality

75 80 74 75 71

When it comes to purchasingproducts, I try to get the verybest or perfect choice

69 68 74 70 78

Getting very good quality is veryimportant to me

69 – 62 53 79

My standards and expectations forproducts I buy are very high

62 – 60 56 75

Factor 3 — recreational and shopping conscious (8%)Going shopping is one of the

enjoyable activities of my life81 73 87 81 80

Shopping the stores wastes mytime*

77 68 85 77 80

I enjoy shopping just for thefun of it

77 75 83 71 –

Shopping is not a pleasantactivity for me*

68 65 69 67 –

I make shopping trips fast* 60 72 – – –

Factor 4 — confused by over choice (7%)There are so many brands to

choose from that I often feelconfused

70 71 62 80 70

Sometimes it's hard to choosewhich stores to shop

67 – 72 57 80

All the information I get ondifferent products confuses me

63 56 70 – 54

The more I learn about products,the harder it seems to choosethe best

62 77 73 – –

Factor 5 — price conscious and value for money (5%)I take time to shop carefully for

the best buys*83 84 76 89 86

I look carefully to find the bestvalue for the money

79 82 69 82 87

I carefully watch how muchI spend*

74 77 73 54 56

Factor 6 — fashion conscious (4%)Fashionable, attractive styling

is very important to me76 79 57 73 76

I keep my wardrobe up-to-datewith changing fashions

73 78 58 78 78

I usually have one or more outfitsof the very newest style

67 72 74 79 65

(continued on next page)

Table 2 (continued)

Consumer styles(variance explained)

Total WestEdmonton

Mall ofAmerica

Pier39

Forumshops

Factor 7 — habitual/brand loyal (4%)Once I find a product or brandI like, I stick with it

71 86 – 77 –

I have favorite brands I buy overand over

70 80 – 72 –

Factor 8 — novelty conscious (3%)It's fun to buy something newand exciting

63 – – – –

To get variety, I shop differentstores and choose differentbrands

60 – – – –

Number 429 121 92 106 110

*Denotes reverse scored items.The columns show factor loadings .60 and higher; decimals are omitted. (Percentagein parenthesis is variance explained for the factor for the total samples).

541S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

Thus, a theory of CDM styles predicts, CDM styles arefound to be independent in structure to different shoppingcontexts.

H2. Demographics and CDM styles

Before examining associations among demographic vari-ables and CDM styles, cluster analysis was run to groupinformants into distinct CDM style orientations. In keepingwith the EDA and comparative method focus of the studyand recognizing the moderate correlations for some of thestyle constructs, the findings include a 30-group clustersolution each with unique combinations of very high tomoderate style orientations—members in each cluster mayhave very high to high scores for more than one CDM style.The aim here is to embrace and understand the complexityoccurring in the data rather than attempting to achieveparsimony by limiting cluster solutions to identifying 5 to 10groups of informants.

Gender is the demographic variable that associates moststrongly with cluster group membership. Fig. 3 shows the resultsfor gender and CDM style groupmembership. Note in Fig. 3 thatthe group high in RFP (i.e., high in recreational conscious,fashion conscious, and perfectionist) includes nearly all females.More males (n=9) than females (n=6) comprise the 0_0_0group (no very high, high, or moderate scores for all styles).

The findings do not include significant direct mainassociations of income, education, and age on CDM styleclusters. For age, this finding supports the Sproles andKendall (1986) conclusion that findings on CDM styles viasamples of high school students are generalizable to other(adult) groups.

H3. Prime reason for visiting an area and CDM styles

Findings (details not shown) from cross-tabulating primereason for visiting an area and CDM style clusters support H3.As enduring personality-like traits, evidence supporting the

Brand consciousness /Price equals quality(α = .70; 4 items)

Fashion conscious(α = .78; 3 items)

.42

Note. All correlations are significant statistically (p < .01).Impulsiveness/careless does not relate significantly with any of the other consumer decision styles.Following Tufte (2003), bold double arrows show moderate size associations (.30 to .5).

Perfectionist / highquality conscious(α = .80; 5 items).35

.28

Recreational & shopping conscious(α = .77; 5 items)

.21.34

NoveltyConscious(α = .45; 2 items)

.25.30

.29

Confused byover choice(α = .69; 4 items)

.21

-.18

Habitual /brand loyal(α = .62; 2 items)

.18

.16

.30

Impulsiveness /Careless(α = .77; 3 items)

Fig. 2. Relationships among consumer decision styles: significant Pearson r correlations.

542 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

alternative versus a null hypothesis for H3 would suggest thatCDM styles are shopping context dependent—the findings donot support such a view.

H4. CDM styles and activities during visit in mall

To probe H4, cluster analyses were run to group informantsby their responses to activities to do and planned during their

0_C_IH

0_FRN_HCP

0_FNP_HB

I_PFCH_0

0_0_0

0_0_B

0_C_0

0_HPB_I

0_CFHB_0

0_BPFCNHR_0

0_C_BNF

0_PHI_NR

0_RCHB_P

P_IBHFCNR_0

0_BCR_IF

0_0_

Cluster Numbe

0

5

10

15

20

25

Bar Char

Key. For example, 0_C_IH group is no style “Very High,” High” Confu

0_ = no “Very High” style C = Confused by over choice F = Fashion conscious R = recreational consciousB = Brand Consciousness

Fig. 3. Gender and consu

current mall visit. A ten cluster solution provides distinctmemberships of single versus multiple-activity groups. SeeTable 3 for the single and multiple activity profiles for the tengroups. Surprisingly and disappointingly, the data analysis doesnot support a direct association between CDM style andactivities group memberships. We speculate that mall typedominates activities done while in a mall across mall visitors no

0

0_C_N

0_0_0

0_P_0

0_HB_FPR

0_NPBFR_0

0_INFR_0

0_CRFH_B

0_0_H

FIR_H_0

0_RFP_0

0_C_HN

0_N_R

0_R_H

0_P_0

r of Case

Demographic1FemaleMale

t

sed about over choice plus moderate Impulsiveness and Habitual

I = Impulsiveness H = HabitualN = Novelty Conscious P = Perfectionist0_0_0 = No “Very High” or “High” style

mer decision-styles.

Table 3Customers' mall activity combinations by planned expenditures

Customer mall activity segment $1–$50 $51–$100 $101–$250 N$250 Total

Exh_Eat_Shb_Soc Count 21 8 19 8 56% within Cluster Number of Case 37.5% 14.3% 33.9% 14.3% 100.0%

Bro_Wal_Eat Count 44 13 6 3 66% within Cluster Number of Case 66.7% 19.7% 9.1% 4.5% 100.0%

Wal_Bro_Shb_Shi_Eat_Soc_Sna_Exh_Con Count 18 17 12 7 54% within Cluster Number of Case 33.3% 31.5% 22.2% 13.0% 100.0%

Con_Bro_Shb_Exh_Eat_Sna Count 16 21 18 10 65% within Cluster Number of Case 24.6% 32.3% 27.7% 15.4% 100.0%

Shb Count 47 26 26 11 110% within Cluster Number of Case 42.7% 23.6% 23.6% 10.0% 100.0%

Soc_Wal_Eat_Shb_Con Count 13 10 6 11 40% within Cluster Number of Case 32.5% 25.0% 15.0% 27.5% 100.0%

Shi_Sna_Shb_Eat Count 13 17 14 9 53% within Cluster Number of Case 24.5% 32.1% 26.4% 17.0% 100.0%

Bro_Shb_Shi_Soc_Eat Count 21 16 15 4 56% within Cluster Number of Case 37.5% 28.6% 26.8% 7.1% 100.0%

Vid_Snc_Eat_Soc_Bro_Exh_Shb Count 13 7 6 5 31% within Cluster Number of Case 41.9% 22.6% 19.4% 16.1% 100.0%

Wal_Shi_Shb_Exh Count 20 4 7 6 37% within Cluster Number of Case 54.1% 10.8% 18.9% 16.2% 100.0%

Total Count 226 139 129 74 568% within Cluster Number of Case 39.8% 24.5% 22.7% 13.0% 100.0%

X2=57.8, df=27; pb .001; ϕ=.32.Dictionary. Bro=browsing; Con=conversation with clerk; Eat=had lunch or dinner; Exh= looked at exhibits; Shb=shopped to buy; Shi=shopped to get ideas;Sna=had a snack; Soc=socialized with friends; Vid=played video games; Wal=walked for exercise.

543S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

matter what their CDM styles; thus, all CDM style clusters areequally likely to attend a movie if a mall offers this option.CDM style may relate more to specific decision making inregard to specific brands of products and services bought ratherthan mall activities.

Table 4Mall activity customer segments by specific mall

Customer mall activities segment WEM

Exh_Eat_Shb_Soc Count 21% within Mall 13.8%

Bro_Wal_Eat Count 11% within Mall 7.2%

Wal_Bro_Shb_Shi_Eat_Soc_Sna_Exh_Con Count 13% within Mall 8.6%

Con_Bro_Shb_Exh_Eat_Sna Count 26% within Mall 17.1

Shb Count 18% within Mall 11.8%

Soc_Wal_Eat_Shb_Con Count 11% within Mall 7.2%

Shi_Sna_Shb_Eat Count 9% within Mall 5.9%

Bro_Shb_Shi_Soc_Eat Count 14% within Mall 9.2%

Vid_Snc_Eat_Soc_Bro_Exh_Shb Count 17% within Mall 11.2

Wal_Shi_Shb_Exh Count 12% within Mall 7.9%

Total Count 152% within Mall 100.0%

X2=102.1, df=27, pb .000; ϕ=.43.Dictionary. Bro=browsing; Con=conversation with clerk; Eat=had lunch or dinneSna=had a snack; Soc=socialized with friends; Vid=played video games; Wal=wa

H5. Planned expenditure level and activities during visit inmall

The findings include a substantial association for plannedexpenditure level and activities during a mall visit (ϕ=.32,

MOA Pier39 Forum shops Total

10 11 14 567.5% 7.2% 9.7% 9.6%32 9 16 6823.9% 5.9% 11.0% 11.6%20 9 13 5514.9% 5.9% 9.0% 9.4%12 12 18 68

% 9.0% 7.8% 12.4% 11.6%27 34 33 11220.1% 22.2% 22.8% 19.2%7 12 12 425.2% 7.8% 8.3% 7.2%4 33 8 543.0% 21.6% 5.5% 9.2%8 18 17 576.0% 11.8% 11.7% 9.8%9 5 3 34

% 6.7% 3.3% 2.1% 5.8%5 10 11 383.7% 6.5% 7.6% 6.5%

134 153 145 584100.0% 100.0% 100.0% 100.0%

r; Exh= looked at exhibits; Shb=shopped to buy; Shi=shopped to get ideas;lked for exercise. Dominating shares by mall appear in bold.

Table 5Consumer decision-making styles and planned expenditures

Cluster number of cases Mean $ Std error of mean N

0_C_IH 100.0 26.9 180_FRN_HCP 135.7 25.2 140_FNP_HB 132.1 46.8 7I_PFCH_0 258.3 116.7 30_0_0 127.8 30.1 180_0_B 87.4 24.2 160_C_0 126.6 24.5 310_HPB_I 110.0 34.2 100_CFHB_0 55.0 12.2 50_BPFCNHR_0 98.8 27.1 210_C_BNF 132.5 27.2 200_PHI_NR 55.0 12.2 50_RCHB_P 65.0 19.4 10P_IBHFCNR_0 225.0 50.0 60_BCR_IF 160.0 49.0 100_0_0 105.0 31.2 150_C_N 116.0 21.1 110_0_0 138.3 35.7 150_P_0 93.4 21.4 190_HB_FPR 115.6 16.0 16O_NPBFR_0 75.0 43.3 80_INFR_0 70.4 17.1 110_CRFH_B 91.7 15.1 180_0_H 58.3 7.1 12FIR_H_0 375.0 0.0 10_RFP_0 95.5 18.0 220_c_HN 85.0 35.6 100_N_R 116.2 31.3 170_R_H 159.6 42.9 130_P_0 86.5 19.7 26Total 110.0 5.5 408

Key. For example, 0_C_IH group is no style “Very High,” High” Confusedabout over choice plus moderate Impulsiveness and Habitual, 0_=no “VeryHigh” style, C=Confused by over choice, I=Impulsiveness, H=Habitual,F=Fashion conscious, R=Recreational conscious, N=Novelty conscious,P=Perfectionist, B=Brand consciousness, 0_0_0=No “Very High” or “High”style.F=1.39, df=29/378, pb .09, η2= .10.

Table 6Planned expenditures in mall by specific mall

Mall Mean $ N Std. error of mean

West Edmonton Mall 129 150 10Mall of America 85 132 9Pier 39 148 147 11Forum Shops 114 139 10Total 120 568 5

F=6.78, df=3/564, pb .000, η2= .04.

544 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

pb .01). Table 3 provides details. The intuitive expectationreceives strong support: engaging in multiple activities thatinclude shopping increases the share of shoppers with plans tospend more than $250. Respondents not engaging in shoppingusually report modest expenditure plans no matter what otheractivities they do engage in during their mall visits.

H6. CDM styles and prime reason for mall visit

The findings do not support H6: CDSM styles are not foundto influence activities during mall visits. One post hoc rationalefor this lack of support is that mall type may dominate the primereason for a mall visit. Given the strong relationship betweenmall type and activities done and planned and support for H7,this post hoc view receives some indirect support.

H7. Prime reason for mall visit and activities done orplanned in the mall

The findings include substantial support for H7. Table 4includes details. Note in Table 4 that for Mall of America closeto 1 in 4 visitors are browsing, walking, and eating (not

shopping)—a share substantially higher than the other threemalls. With respect to nomological validity, such a findingwould indicate that lower overall planned expenditures forMOA visitors compared to visitors to the other three malls (anobservation that relates to H13; this findings support thisspeculation (see Table 6 for details).

H8. Activities and satisfaction with mall

The findings do not support H8. Distinct patterns of activitiesdone and planned do not relate to visitors' reports of specificand global measures of mall satisfaction. Apparently, browsingand not buying leads to a comparable level of satisfaction asshopping. The quality of the activity experienced and not thecombination of activities done or planned may affect satisfac-tion with mall attributes and malls globally.

H9. Global mall satisfaction and intention to visit mall

The findings includes a substantial positive associationbetween global mall satisfaction and intention to “come back tothe city specifically to visit the mall” (ϕ=.33, pb .000). Giventhe widespread support of this hypothesis in the retail serviceliterature, a finding other than substantial positive supportwould have indicated low nomological validity for the modelthat Fig. 1 depicts.

H10. Mall type and mall activities

Mall visitors can engage only in activities available to do inthe mall. Thus, intuitively the expectation is for substantialsupport to be found for H10. The findings from the data analysissupports H10 (ϕ=.42, pb .001). See Table 4 for details.

H11. CDM styles and planned expenditure level

The findings provide support for H11. For comparing meanplanned expenditures by CDM style clusters, the midpoint valuewas assigned to each of the first three expenditure rangesprovided in the survey for responding to the plannedexpenditure level (an arbitrary value of $400 was assigned forresponses checking the above $250 level). Using this approachthe overall mean planned expenditure is $110. The one grouphaving a very high perfectionist style had an average plannedexpenditure equal to $225 (n=6). The group with the lowestaverage planned expenditure level (mean=$55, n=5) does notinclude a very high, high, or moderate perfectionist style. Table5 includes details. The findings from a cross-tabulation analysissupport the same conclusion.

$ 30 − − 60 > $ 60> $ 30 Income (000)

$ 150

140

130

120

110

100

90

80

70

116114

148 Female

118

135 Male

75

Planned Expenditure Level

Note. Test for linearity of income influence on planned expenditures among female informants: F = 4.09, d.f. = 1/328, p < .04; eta2 = .02. Test for linearity of income influence on planned expenditures among male informants: F = 8.42, d.f. = 1/183, p < .004; eta2 = .05.

Fig. 4. Gender and income influence on planned expenditure level (Today in Mall).

545S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

The linkage between expenditures level and shopping mayhave greater relevance for brand choice rather than for mall visit.Certainly, visitors very high in price consciousness may beexpected to plan to spend less per brand than visitors very high inperfectionism and brand consciousness but the present studydoes not focus on brand buying behavior.

Mall V

Female

< 30 > 60 30 − 60

C

C$75 (6)

$175 (2)

P+

$150 (4)

F+

$113 (4)

$75 (1)

P+

$88 (4)

F+

$200 (2)

C

$275 (2)

P+ $188 (4)

F+

Note. Absent branch indicates zero cases (e.g., no Cmales, < $30,000 annual income among respondents).

Key. C = Confused by overchoice P+ = Perfectionist at highest level plus impulsiveness,

brand consciousness, habitual, fashion consciousness, confused by overchoice, novelty,recreational conscious

F+ = Fashion conscious, recreational conscious, and novelty conscious

HB = Habitual and Brand Consciousness

Income $ 000’s

HB

$142 (3)$138 (4)

HB

$125 (4)

HB

Fig. 5. How gender, income, decision sty

H12. Planned expenditure level and global mall satisfaction

The findings support a moderate (r= .18) positive relation-ship between planned mall expenditure level and global mallsatisfaction. The dictum that shopping is fun when you plan tospend money bears up.

isitor

Male

< 30 30 − 60 > 60

$75 (3)

C

$375 (1)

P+

$125 (2)

C

F +

$25 (1)

$25 (1)

$50 (2)

HB

$125 (2)

HB

HB

les influence planned spending (n).

Table 7Consumer decision style groups and overall satisfaction with shopping mall

Cluster number of cases Mean Std error of mean N

0_C_IH 13.0 .53 170_FRN_HCP 13.7 .39 130_FNP_HB 14.1 .34 7I_PFCH_0 12.7 1.20 30_0_0 12.0 .47 180_0_B 11.9 .58 140_C_0 11.3 .58 300_HPB_I 12.4 .67 90_CFHB_0 12.6 .93 50_BPFCNHR_0 13.6 .41 180_C_BNF 12.8 .48 200_PHI_NR 13.2 1.28 60_RCHB_P 13.4 .45 10P_IBHFCNR_0 14.7 .21 60_BCR_IF 13.2 .36 90_0_0 11.1 .77 160_C_N 12.8 .59 80_0_0 12.5 .50 150_P_0 12.5 .46 170_HB_FPR 13.1 .60 15O_NPBFR_0 12.8 .62 80_INFR_0 12.6 .86 100_CRFH_B 13.4 .33 170_0_H 12.3 .73 10FIR_H_0 12.5 2.50 20_RFP_0 13.5 .41 220_c_HN 11.8 .80 90_N_R 13.2 .39 170_R_H 13.7 .47 110_P_0 12.0 .55 25Total 12.7 .11 387

F=1.86, df=29/357, pb .005, η2= .13.

546 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

H13. Mall type and planned expenditure levels

Some mall types associate with higher planned expenditurelevels than alternative mall types: the findings support H13. SeeTable 6 for details.

The findings in Tables 3 and 6 support the nomologicalvalidity of the study. For example, the very low averageplanned expenditure for MOA visitors should and doesmatch with a very high proportion of MOA visitors engagingin activities other than shopping (e.g., browsing, walking,and eating: 24%), while the opposite pattern occurs for Pier39 visitors.

H14. Demographics and planned expenditure levels

Supporting H14: the association is substantial betweenannual income before taxes and planned expenditure levelsboth for female and male visitors but especially for malevisitors. Fig. 4 provides details. The average overall plannedexpenditure by males is lower than females due to the verysubstantial gender by income interaction effect: male visitorswith low incomes have a very low average expenditure level.

H15. Demographics moderates the influence of CDM styleson planned expenditures

The findings support H15. Demographics moderate theinfluence of consumer decision-making styles on plannedexpenditures. Among both females and males high incomeenables consumers highly perfectionist and brand conscious toplan high expenditures; female, high income, high perfectionistare more likely to plan high expenditures versus male, highincome, high perfectionist—and visitors with both of theseprofiles have higher planned expenditures than other mallvisitors. Worth emphasizing here is that while intuitivelyappealing, this hypothesis came-to-mind only after examiningthe data and after failing to find support for H4.

Fig. 5 is a tree diagram for gender by income by CDM Styleand average mean planned expenditure levels. These resultsinclude the informants with the highest average spikes inplanned expenditure levels.

Fig. 5 indicates that among females the perfectionist (P) stylein combination of high income relates with the very high levelof average expenditure. However, only two informants fit thiscombination. Note in Fig. 5 that the average expenditure for alltree branches that include habitual and brand consciousness(HB) have average expenditures below $145 while three of thefour P style branches have average expenditures above $170.Conclusion: planned expenditures per shopping trip to mallsvary substantially by specific combinations of gender, income,and CDM styles.

H18. CDM styles and global satisfaction with mall

Because the present study examines learning styles theoret-ically but not empirically, the analyses does not include findingsfor H16 and H17. The findings do support H18: CDM stylesinfluence satisfaction with mall shopping.

The results in Table 7 follow from the single-item scale forglobal satisfaction. The group of informants with all averageand below average normalized CDM style scores (O_O_O) hasthe lowest average global satisfaction score (M=11.1, se= .77)and the group very high in P has the highest average globalsatisfaction score (M=174.7, se= .21). See Table 7 for completedetails.

Fig. 6 provides details for P for global mall satisfaction forboth genders and for each of the four malls. The patterns areconsistent: the average level of satisfaction is highest forinformants in the highest versus the lowest P style quintile forboth genders and all four shopping malls. The findings are thesame for the other CDM styles (details available from theauthors).

5. Conclusions and implications for theory and retailingstrategy

The main conclusion is that empirical research does supportthe existence of CDM styles among adult mall shoppers indifferent mall contexts. Gender is a prime antecedent associ-ating with CDM styles. The influence of CDM styles on mallshopping consequences is subtle and indirectly influencesactivities during mall visits via influencing planned expenditurelevels. Consumers very high in perfectionism are most likely to

14

13

12

11

10

1 2 3 4 5 1 2 3 4 5

Mall Satisfaction

Perfectionist Consumer DM Style Level

14

13

12

11

10

Mall Satisfaction

Perfectionist Consumer DM Style Level

Females

Males

West EdmontonMall

Mall ofAmerica

Pier 39

Forum Shops

Note. Compare level 5 (higher in satisfaction)with level 1 perfectionists for shoppersin Mall of America (same for other 3 malls).

Note. Average mall satisfaction highest for level 5compared to level 1 perfectionists for both femalesand males (p < .001).

Fig. 6. Perfectionist consumer decision-making style and overall satisfaction with mall.

547S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

high in planned mall expenditures. CDM styles associatesubstantially with visitor satisfaction with visiting shoppingmalls.

Further work on CDM styles needs to further addressthe study of antecedents and consequences and to furtherclarify the variety of types of consumers by multiple styleorientations.

Mall and retail store strategists may find that communicatingto shoppers by certain shoppers' CDM style orientations islikely to be effective in increasing their shoppers' plannedexpenditures levels—not by changing the expenditures plannedbut by changing the clientele toward customers planning highexpenditures.

6. Limitations and implications for further research

This article reports an EDA for expanding the research viewof CDM styles and antecedents and consequences of suchorientations. Limitations include focusing only on CDM stylesamong mall visitors and the attempt to consider only how mallvisit behavior relates to CDM styles. The findings do not focuson brand choice behavior of shoppers. The antecedents andconsequences of CDM styles within an empirical researchprogram focusing on brand choice is the focus that might be bestto plan for further research.

Acknowledgment

The authors gratefully acknowledge funding of this study bya grant from the International Council of Shopping CentersEducational Foundation.

References

Behrens J. Principles and procedures of exploratory data analysis. PsycholMethods 1997;2(2):131–60.

Bellenger DN, Korgaonker PK. Profiling the recreational shopper. J Retail1980;56(3):77–91.

Bettman JR, Zins MA. Constructive processes in consumer choice. J ConsumRes 1977;4:75–85 (September).

Cook T, Campbell D. Quasi-experimentation: design and analysis issues forfield settings. Chicago: Rand-McNally; 1979.

Cronbach L, Meehl P. Construct validity in psychological tests. Psychol Bull1955;52:281–302.

Darden WR, Reynolds FD. Shopping orientations and product usage rates.J Mark Res 1971;8:505–8 (November).

Deighton J. The consumption of performance. J Consum Res 1992;19:362–72(December).

Fabrigar L, Wegener D, MacCallum R, Strahan E. Evaluating the issue ofexploratory factor analysis in psychological research. Psychol Methods1999;4(3):272–99.

“Fast Facts”. West Edmonton Mall 2002 (March): from the World Wide Web:http://westedmontonmall.com/info/fastfacts.html.

Getz D. Tourist shopping villages. Tour Manage 1993:15–26.Ginsburg J. Xtreme retailing: stores fight the online onslaught. Business week

(December 20): cover story; 1999.Gladwell M. The tipping point. New York: Little Brown; 2000.Guralnik DB, editor. Webster's new world dictionary (2nd college ed.). Akron,

OH: William Collins; 1976.Hiu A, Siu N, Wang C, Chang L. An investigation of decision-making styles of

consumers in China. J Consum Aff 2001;35(2):326–45.Howard J, Sheth J. The theory of buyer behavior. New York: Wiley; 1969.Kamaruddin AR, Mokhlis S. Consumer socialization, social structural factors

and decision-making styles: a case of adolescents in Malaysia. Int J ConsumStud 2003;27(2):145–57.

Joos SK, Hickam DH, Borders LM. Patients' desires and satisfaction in generalmedicine clinics. Public Health Rep 1993;108(6):751–9.

Lisanti, T. “A mega mall for the new millennium,” Discount Store News 1998(September 7): from the World Wide Web: http://findarticles.com/cf_0/m3092/n17_v37/21122322/print.jhtml.

“Mall of America continues its nine-year reign as nation's largest mall.” PRNewswire 2001 (August 9): from the World Wide Web: http://findarticles.com/cf_0/m4PRN/2001_August_9?77089383/print.jhtml.

McClelland D. Identifying competencies with behavioral-event interviews.Psychol Sci 1998;9(5):331–9.

McDonald WJ. The role of demographics, purchase histories and shopperdecision-making styles in predicting consumer catalogue loyalty. J DirectMarket 1993;7(3):55–65.

Martin J. Real perspective selves. Theory Psychol 2005;15(2):207–14.

548 S. Wesley et al. / Journal of Business Research 59 (2006) 535–548

Oliver RL, Rust RT, Varki S. Customer delight: foundations, findings, andmanagerial insight. J Retail 1997;73:311–36.

Peter JP. Construct validity: a review of basic issues and marketing practices. JMarket Res 1981;18:133–45 (May).

“Pier 39 press priorities.” Pier 39 (March): from the World Wide Web: http://www.pier39.com/pier30/english/press/index.html.

Ragin C. The comparative method: moving beyond qualitative and quantitativestrategies. Berkeley: University of California Press; 1987.

Tai SHC. Shopping styles of working Chinese females. J Retail Consum Serv2005;12:191–203.

Tufte E. The visual display of quantitative information. Cheshire, CT: GraphicPress; 2003.

Tukey J. Exploratory data analysis. Reading, MA: Addison-Wesley; 1977.Sears DO. College sophomores in the laboratory: influences on a narrow data

base on social psychology's view of human nature. J Pers Soc Psychol1986;51(3):515–30.

Sharma Arun, Levy Michael. Categorization of customers by retail salespeople.J Retail 1995;71(1):71–81.

Sproles GB. From perfectionism to dadaism: measuring consumers' decision-making styles. Proceedings American Council on Consumer InterestConference, Columbus, OH; 1985. p. 79–85.

Sproles GB, Kendall EL. A methodology for profiling consumers' decision-making styles. J Consum Aff 1986;2(2):267–79.

Sproles EK, Sproles GB. Consumer decision-making styles as a function ofindividual learning styles. J Consum Aff 1990;24(1):134–47.

Sproles EK, Cox DE, Sproles GB. Characterizing vocational students' learningstyles: a replication and generalization of findings. J Voc Educ Res 1987;12(4):1–11.

Sternberg RJ, Grigorenko E. Are cognitive styles still in style? Am Psychol1997;52(7):700–12.

Stone GP. City shoppers and urban identification: observations on the socialpsychology of city life. Am J Sociol 1954;60:36–45.

Walsh G, Hennig-Thurau T, Wayne-Mitchell V, Wiedmann K-P. Consumers'decision-making style as a basis for market segmentation. J Target MeasAnal Mark 2001a;10(2):117–31.

Walsh G, Mitchell V-W, Hennig-Thurau T. German consumer decision-makingstyles. J Consum Aff 2001b;35(1):73–81.

Wang C-L, Siu N, Hui A. Consumer decision-making styles on domestic andimported brand clothing. Eur J Mark 2004;38(1/2):239–52.

Wells WD. Discovery-oriented consumer research. J Consum Res 1993;19:489-504.

Weick K. Sensemaking in organizations. Thousand Oaks, CA: Sage; 1995.Weiss MJ. To be about to be. Am Demogr 2003;25:29–36 (September).Wiggins M. X-it plans. Am Demogr 2004;26:35–8 (May).Westbrook RA, Black W. A motivation shopper based typology. J Retail

1985;61:78–103 (Spring).