518539

7/27/2019 518539

1/13

Journal of Consumer Research, Inc.

My Brand or Our Brand: The Effects of Brand Relationship Dimensions and SelfConstrualon Brand EvaluationsAuthor(s): Vanitha Swaminathan, Karen L. Page, and Zeynep GrhanCanliSource: Journal of Consumer Research, Vol. 34, No. 2 (August 2007), pp. 248-259Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/10.1086/518539.

Accessed: 29/10/2013 14:05

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at.http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago PressandJournal of Consumer Research, Inc.are collaborating with JSTOR to

digitize, preserve and extend access toJournal of Consumer Research.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=ucpresshttp://www.jstor.org/stable/10.1086/518539?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/10.1086/518539?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=ucpress

7/27/2019 518539

2/13

248

2007 by JOURNAL OF CONSUMER RESEARCH, Inc. Vol. 34 August 2007

All rights reserved. 0093-5301/2007/3402-0011$10.00

My Brand or Our Brand: The Effects ofBrand Relationship Dimensions and Self-

Construal on Brand Evaluations

VANITHA SWAMINATHANKAREN L. PAGEZEYNEP GURHAN-CANLI*

Consumer-brand relationships can be formed based on individual- or group-levelconnections. For example, a consumers relationship with a Mercedes may bebased on the desire to express individual-level unique identity (e.g., self-conceptconnection), whereas a relationship with a local brand (e.g., Ford) may be basedon a group-level patriotic national identity (e.g., country-of-origin connection). We

suggest that the effects of self-concept connection and brand country-of-originconnection vary basedon self-construal.Results across two studies reveal that,underindependent self-construal, self-concept connection is more important. Under inter-dependent self-construal, brand country-of-origin connection is more important.

R ecently, the relationship perspective has become in-creasingly popular as a theoretical lens for understand-ing consumer-brand interactions (Aaker, Fournier, and Bra-sel 2004; Escalas and Bettman 2005; Fournier 1998).Consumers are known to form strong relationships withthose brands that have values and personality associationsthat are congruent with their self-concept (Sirgy 1982). Inthis way, brand relationships can be viewed as expressionsof consumers identities (Escalas and Bettman 2005; Reed2004).

Self-concept connection, a dimension of the consumer-brand relationship, indicates the amount that the brand con-tributes to ones identity, values, and goals (Fournier 1998).For instance, the Harley Davidson brand, with its free-spir-ited and rebellious image, is likely to appeal more to those

*Vanitha Swaminathan is assistant professor of marketing, Universityof Pittsburgh, Pittsburgh, PA 15260 ([email protected]). Karen L. Pageis assistant professor of marketing, Texas A&M University, College Station,TX 77843-4112 ([email protected]). Zeynep Gurhan-Canli is associateprofessor of marketing, Koc University, Istanbul, Turkey ([email protected]

.tr). Correspondence: Vanitha Swaminathan. The authors gratefully ac-knowledge a research grant from the International Business Center at theKatz Graduate School of Business and helpful input in reviews from theeditor, associate editor, and reviewers. In addition, the authors thank RohiniAhluwalia, Esra Gencturk, Andrew Gershoff, Jeff Inman, Chris Jani-szewski, Vikas Mittal, Aysegul Ozsomer, Serdar Sayman, Stijn Van Os-selaer, and David Wooten for their helpful comments on previous versions.

John Deighton served as editor and Susan Broniarczyk served as associateeditor for this article.

Electronically published May 24, 2007

individuals whose self-concept contains these traits. There-fore, a high self-concept connection can symbolize a con-sumers individual identity.

Another stream of research suggests that brand relation-ships can furnish participants with a social identity (Weiss1974; Wright 1974). Consistent with this idea, consumer

research has shown that brands can be used to communicateand reinforce national identity (Johansson 1989; Shimp andSharma 1987). Further, brand attitudes have been shown tovary based on country of origin (Gurhan-Canli and Ma-heswaran 2000; Hong and Wyer 1990; Maheswaran 1994).For example, when France refused to participate in the U.S.-led Iraq war, Americans demonstrated their displeasure byboycotting French-made wines and increasing their pur-chases of U.S.-made wines (Chavis and Leslie 2005). Thisexample highlights the notion that brand country-of-originconnection can form an important facet of a consumersrelationship with a brand.

The research to date has primarily focused either on self-concept connection (Escalas and Bettman 2005; Fournier

1998) or brand country-of-origin connection (Gurhan-Canliand Maheswaran 2000). Synthesizing these two researchstreams, we examine the differential role of self-conceptconnection and brand country-of-origin connection withina unified conceptual framework. The unified framework al-lows us to examine when and how these dimensions ofconsumer-brand relationships are more relevant or salientin influencing brand evaluations.

Consumers have been shown to express different aspectsof the self under varying circumstances (Reed 2004). In-

This content downloaded from 130.239.76.10 on Tue, 29 Oct 2013 14:05:24 PMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp

7/27/2019 518539

3/13

BRAND RELATIONSHIPS AND SELF-CONSTRUAL 249

dividuals with an independent self-construal view them-selves as separate individuals, whereas individuals with aninterdependent self-construal view themselves as part of agroup (Markus and Kitayama 1991). Further, research sug-gests that most individuals have a dynamic self consistingof both independent and interdependent traits and situational

cues can activate either type (Agrawal and Maheswaran2005; Triandis 1995). Drawing on this perspective of thedynamic self, it is suggested here that the impact of self-concept connection (with its focus on the individual) andbrand country-of-origin connection (with its focus on thegroup) may vary based on an individuals construal of self.

The context for examining these research issues is brandattitude change following negative brand information con-cerning a closely related line extension. Negative infor-mation or negative publicity surrounding a brand canthreaten the stability of the consumer-brand relationship andhas a higher salience and diagnostic value than positiveinformation (Aaker et al. 2004). According to research onbranding, a key benefit of strong consumer-brand relation-

ships is their ability to help maintain brand attitudes in theface of negative information (Ahluwalia, Burnkrant, andUnnava 2000; Fournier 1998). Because brand attitudechange is likely to vary significantly based on relationshipstrength, it provides an ideal context for investigating therelative importance of individual- and group-based con-sumer-brand relationship dimensions.

In summary, the extent to which the negative brand in-formation is likely to influence brand attitudes will dependon (a) self-concept connection, (b) the brand country-of-origin connection, and (c) self-construal. In study 1, wedemonstrate how self-construal moderates the effect of bothself-concept connection and brand country of origin. Study2 replicates this effect in a different context (i.e., athletic

shoes).

THEORETICAL BACKGROUND

Based on social identity theory (Tajfel and Turner 1979),the self can be conceptualized as composed of a personalidentity and a group identity. Object ownership has beenshown to contribute to ones unique individual identity (Belk1988). Material possessions can also help communicategroup identity. Wallendorf and Arnould (1988, 532) dem-onstrate that attachments to objects are signs of ones con-nection to or differentiation from other members of society.

Extending identity theory to the realm of consumer be-havior, Kleine, Kleine, and Allen (1995) conceptualize ma-

terial possession attachment as having two distinct fac-etsone facet reflects consumers desire for a uniquepersonal identity (i.e., autonomy seeking), and a secondfacet reflects a desire for group identity (i.e., affiliation seek-ing). Applying the theory of material possession attachmentto the context of brand relationships, it is suggested herethat unique self-concept connection is based on the need forindividual autonomy and brand country-of-origin connec-tion is based on a need for affiliation with a group. Weelaborate on these ideas next.

Self-Concept Connection

Recall that the self-concept connection reflects the degreeto which the brand is used to express a significant aspectof the individual self (Fournier 1998). While self-conceptconnection can be based on both personal and group iden-

tities (as per social identity theory; Tajfel and Turner 1979),we distinguish between personal and group aspects and con-ceptualize self-concept connection as the relationship be-tween a consumer and a brand on the basis of a connectionbetween a consumers unique self and what the brand sym-bolizes for the consumer. In other words, self-concept con-nection focuses on personal identity and not on group iden-tity. In many independent cultures (where Fourniersresearch was conducted), an individuals self-concept is pri-marily unique, abstracted from the social environment, andindependent of others.

What are some of the benefits of a strong self-conceptconnection? A strong consumer-brand relationship is be-lieved to encourage relationship durability and greater tol-

erance when the brand perception is challenged by negativecircumstances (Ahluwalia et al. 2000; Fournier 1994, 1998).Consequently, we expect that consumers with a high uniqueself-concept connection, a dimension of consumer-brand re-lationships, will be more likely to disregard the negativebrand information. In support of this, Pomerantz, Chaikenand Tordesillas (1995) show that when attitudes are viewedas central to the self-concept, individuals are more likely todemonstrate selective memory and resistance to counter-attitudinal information.

Self-Construal

A growing body of work suggests that the self is a com-

plex structure with multiple manifestations (Singelis 1994).An individuals construal of the self is frequently definedas a constellation of thoughts, feelings, and actions con-cerning ones relationship to others such as the self beingdistinct from others or connected to others (Singelis 1994,581). While self-construal is often considered to be basedon cultural orientation (Markus and Kitayama 1991), re-search has found that self-construal can be activated throughsituational priming (Agrawal and Maheswaran 2005; Ng andHouston 2006; Triandis 1995).

According to Markus and Kitayama (1991, 226), the in-dependent self-construal is characterized by an individualwhose behavior is organized and made meaningful primarilyby reference to ones own internal repertoire of thoughts,

feelings, and actions, rather than by reference to thethoughts, feelings, and actions of others. This implies thatan independent self-construal should cause ones opinionsto be altered by individual thoughts and personal opinions.Conversely, those with an interdependent construal of selfbase their attitudes and behavior on the thoughts, feelings,and actions of others in the relationship (Markus and Ki-tayama 1991).

Different aspects of the self have been shown to influenceconsumer behavior at different points in time (Reed 2004).


7/27/2019 518539

4/13

250 JOURNAL OF CONSUMER RESEARCH

Consumer behavior research has shown that individualsconstrual of self (independent or interdependent) can influ-ence brand meaning (Escalas and Bettman 2005), the per-suasiveness of various advertising appeals (Agrawal andMaheswaran 2005), and brand extension evaluations (Ngand Houston 2006). We suggest that a consumers self-con-

cept connection is likely to be important only when an in-dependent self-construal becomes salient. The self-conceptconnection is based on consumers desires to express theirindividuality and their self as distinct from others. Therefore,the impact of self-concept connection is likely to be greaterwhen an independent self-construal is primed.

On the one hand, if an independent self-construal is rel-atively salient and there is a high self-concept connectionwith the brand, the individual is more likely to refute anynegative information challenging his or her self-conceptconnection with the brand by counterarguing the negativeinformation (Ahluwalia et al. 2000). Such counterargumentswill minimize any brand attitude change caused by negativeinformation. However, if an independent self-construal is

relatively salient and there is a low self-concept connectionwith the brand, then brand attitude change will occur, sincethere is no refutation of negative information.

On the other hand, when an interdependent self-construalis made salient, self-concept connection is likely to be lessimportant when evaluating a brand. For instance, Markusand Kitayama (1991, 236) suggest that, among those withmore interdependent selves, ones inner feelings may be lessimportant in determining ones consequent actions.

The preceding arguments imply that the impact of self-concept connection should vary based on self-construal,such that changes in brand attitude should be greater forlow self-concept connection (vs. high self-concept connec-tion) brands under independent self-construal. The preced-

ing arguments also suggest that, when an interdependentself-construal is made salient, brand attitude change willoccur regardless of whether the individual has a high or lowindividual-based relationship with the brand. This leads tothe following hypothesis:

H1a: The impact of self-concept connection on theextent of brand attitude change varies based onself-construal. When an independent self-con-strual is primed, exposure to negative infor-mation about the brand leads to lower brandattitude in the negative information condition(relative to a control group) for consumers hav-ing lower self-concept connection. For consum-

ers having higher self-concept connection, therewill be no change in brand attitude in the neg-ative information condition (relative to a controlgroup).

Brand Country-of-Origin Connection

Brand country-of-origin connection is the extent to whicha brand is used to express ones patriotic national identity.

Researchers have demonstrated that country of origin hasan important impact on consumer evaluations of productsand brands (Hong and Wyer 1990; Maheswaran 1994).Shimp and Sharma (1987) found that ethnocentric individ-uals were more likely to purchase domestic products overforeign products, reinforcing their patriotic identity. Simi-

larly, Klein, Ettenson, and Morris (1998) found that evenwhen a consumer recognizes a brand as high quality, theconsumer may still refuse to purchase the brand if it doesnot adequately symbolize their group membership. Fur-thermore, John and Klein (2003) suggest that consumersoften participate in boycotting specific brands (e.g., religiousorganizations boycotting Disney) because it strengthenstheir group identity.

The impact of country of origin on brand evaluations isshown to vary based on a variety of factors, including priorelaboration (Hong and Wyer 1990), consumer expertise(Maheswaran 1994), consumer ethnocentricity (Shimp andSharma 1987), and culture-specific factors (Klein et al.1998). It is suggested here that the brand country of origin

(i.e., whether the brand name is local or foreign), a grouplevel variable, is particularly meaningful to consumers as ithelps differentiate between in-group members and out-groupmembers. Further, brand country-of-origin connection maybecome more prominent or salient when there is a greaterfocus on relationships with others. Recall that an interde-pendent self-construal places greater emphasis on the rela-tionship between self and others, with distinctions madebetween in-group and out-group members. Because of this,brand country of origin, which distinguishes between localand foreign brands, is likely to have a greater impact whenself-construal is relatively more interdependent. Specifically,consumers are more likely to resist negative informationregarding local (vs. foreign) brands when an interdependent

self-construal is primed. In contrast, since an independentself-construal places greater emphasis on self, brand countryof origin is expected to have less influence on brand attitudeswhen an independent self-construal is primed. In support ofthis, Gurhan-Canli and Maheswaran (2000) show that homecountry products are evaluated more favorably when theinterdependent self-construal is dominant.

The preceding arguments imply that the impact of brandcountry-of-origin connection should vary based on self-con-strual, such that changes in brand attitude should be greaterfor foreign (vs. local) brands under interdependent self-con-strual. However, the impact of brand country-of-origin con-nection will be lower in independent self-construal. Thisleads to the following hypothesis:

H1b: The impact of brand country-of-origin connec-tion on the extent of brand attitude change var-ies based on self-construal. When the brand isof foreign origin and an interdependent self-construal is primed, there should be a lowerbrand attitude in the negative information con-dition (relative to a control group). When thebrand is of local origin and an interdependentself-construal is primed, there should be no


7/27/2019 518539

5/13


change in brand attitude in the negative infor-mation condition (relative to a control group).Brand attitudes do not vary as a function ofbrand country-of-origin connection when an in-dependent self-construal is primed.

STUDY 1

Our methodological approach for testing the hypothesesinvolved using real brand names. Further, the empirical con-text had to be such that the category represented a highdegree of risk to the consumer, so that negative brand in-formation could potentially have an impact on brand atti-tudes. Based on this criterion, consumer electronics was thechosen product category. To manipulate negative brand in-formation, the context for the study was an electronics brandintroducing a failed line extension.

Pretests. To study the effect of self-concept connectionand brand country of origin on brand attitudes followingexposure to negative information, we chose various con-sumer electronics brands. Further, to test the impact of vary-ing levels of self-concept connection (which are measuredfor each respondent), the brand names chosen had to rep-resent a broad range of self-concept connection scores, rang-ing from very low to very high. In addition, since brandcountry-of-origin connection was to be manipulated basedon whether brands were of local origin or foreign origin,various pairs of local and foreign brands were selected forfurther pretesting.

In the pretest ( ), various local and foreign elec-n p 42tronics brands were presented to respondents who were

asked to rate these brands in terms of familiarity (1p highlyfamiliar; 5 p highly unfamiliar) and brand attitude, as mea-sured on four five-point scales (low quality/high quality,bad/good, unfavorable/favorable, negative/positive). The fourbrand attitude scores were then averaged to form an eval-uation index ( ). Based on subject responses, thisap .95pretest yielded Samsung and Dell as two consumer elec-tronics brands with both high familiarity ( ,M p 1.62Samsung

; , NS) and similar levels of initialM p 1.51 t(41) p 0.50Dellbrand attitude ( , ;M p 3.84 M p 4.01 t(41) pSamsung Dell

, NS).1.21In a second pretest ( ), in order to eliminate pos-n p 42

sible alternative explanations (e.g., brand breadth and per-ceived fit), we determined the brand breadth of the Samsung

and Dell brand names by having respondents list all productcategories that came to mind when presented with the Sam-sung/Dell brand names. The number of product categoriesassociated with a given brand name was computed. BothSamsung and Dell exhibited similar brand breadth( , ; , NS). Addi-M p 3.40 M p 3.52 t(41) p 0.32Samsung Delltionally, the perceived fit of each brand with televisions wasexamined. Participants responded to three five-point items(extremely bad fit/extremely good fit, not at all logical/ex-tremely logical, not at all appropriate/extremely appropriate)

that were averaged into a perceived fit score ( ). Theap .92perceived fits of Samsung and Dell with a television weresimilar ( , ; , NS),M p 3.4 M p 3.5 t(41) p 0.78Samsung Delland both brands currently manufacture televisions.

To determine the extent of brand relationship strength(i.e., self-concept connection and the brand country-of-or-

igin connection) of the two brands, respondents rated Delland Samsung on both self-concept connection and brandcountry-of-origin connection. Self-concept connection withthe brand was measured through five statements taken froma brand relationship quality scale developed by Fournier(1994). The statements are as follows: The brand and Ihave a lot in common, This brands image and my selfimage are similar in a lot of ways, This brand says a lotabout the kind of person I am or want to be, This brandreminds me of who I am, and This brand is a part of me.Responses to each statement were measured on a five-pointscale (1 p strongly agree; 5 p strongly disagree) and wereaveraged to obtain a self-concept connection score (ap

). Both Samsung and Dell had similar levels of self-.93

concept connection ( , ;M p 2.20 SD p 1.1 M pSamsung Dell, ; , NS). Additionally, there was2.30 SD p 1.1 t(41) p 0.58

a broad distribution of self-concept connection scoresaround this average ranging from very high to very low.

To judge the degree to which Dell was representative ofa local brand and Samsung represented a foreign brand, thesame pretest measured the extent of brand country-of-originconnection of each brand. For instance, participants re-sponded to the following statements: I associate the Dell(Samsung) brand name with things that are American (for-eign), To me, Dell (Samsung) brand represents whatAmerica (a foreign country) is all about. Responses wereindicated on a five-point scale (1 p strongly disagree; 5 pstrongly agree) and averaged to determine the brand coun-

try-of-origin connection of the Dell brand ( ) andap .89Samsung brand ( ). The results of the pretest con-ap .95firmed that Dell and Samsung were significantly differentin terms of their American-ness (Dell p 3.87; Samsungp 1.82; , ) and in terms of their for-t(41) p 9.58 p !.01eign-ness (Dell p 1.71; Samsung p 3.04; ,t(41) p 6.66

).p !.01

Procedure. Participants received a questionnaire andwere informed that they could voluntarily participate in thismarketing study about products and brands and receive anincentive of either $3 cash or course credit in return fortheir participation. The questionnaire first asked students toindicate their familiarity with and use of the brand and then

completed the brand relationship scale using the scale itemsdescribed previously. Measures of self-concept connectionelicited from respondents were used in the subsequent anal-ysis. This was followed by a filler task to clear short-termmemory that asked students to evaluate an unrelated brand.Next, the self-construal prime was administered, followedby negative brand information (or no information in the noinformation condition) and participants evaluation of thebrand. Finally, the students completed the self-construal ma-nipulation check and finished with a brief demographic sec-


7/27/2019 518539

6/13


tion. After the study, participants responded to an open-ended suspicion probe.

Independent Variables

Self-Concept Connection. As described previously,

self-concept connection was measured using five items fromFourniers (1994) brand relationship quality scale. Responsesto each statement were measured on a five-point scale (1 pstrongly agree; 5 p strongly disagree) and were averagedto obtain a self-concept connection score ( ).ap .93

Self-Construal. Self-construal was primed using Trafi-mow, Triandis, and Gotos (1991) method. Participants wereinstructed to take 5 minutes to think about and write downhow they are similar to (interdependent) or different from(independent) their friends and family.

Design

A total of 320 students (50% male; 99% ages 1824) atthe University of Pittsburgh participated in the study. Noneof the students were of foreign origin, and all of the studentsindicated at least moderate familiarity (less than or equal tothree on a five-point scale; 1 p highly familiar; 5 p highlyunfamiliar) with the brand names in the study. They wererandomly assigned to conditions in a 2 # 2 # 2 (infor-mation valence: negative vs. no information [manipulated],self-construal: interdependent or independent [manipulated],brand country of origin: local vs. foreign [manipulated])between-subjects design. Self-concept connection was mea-sured as described earlier.

Negative Information. To create a negative informa-tion scenario, we presented information about a fictitiousline extension and hypothetical ratings on three attributesof the extension, along with the competitors ratings on thesame attributes. Participants read the following information:We are interested in obtaining your evaluations with regardto a new television that Samsung has introduced. Followingare the ratings of Samsungs new line of Super Fine 34-inch Model WS34V1 by an independent product testingagency. This line has recently been introduced in the marketplace. Please read carefully through the ratings of Samsungsnew line of televisions as well as its competitors and answerthe questions that follow. The ratings for the new Samsungtelevision and its competitors televisions (Sony and GE)were presented to the participants in a table. In the hypo-thetical ratings (1 p very high; 7 p very low), the Samsungtelevision received a seven on picture quality, seven onsound clarity, and heavy on weight. The Sony competingtelevision received a two on picture quality, two on soundclarity, and light on weight. The GE competing televisionreceived a four on picture quality, four on sound clarity, andlight on weight.

Brand Attitude. After reading the negative line exten-sion ratings, participants evaluated the brand name in generalaccording to the same four brand attitude items used in the

pretest, each measured on a five-point scale and then av-eraged to form an evaluation index ( ).ap .95

Results

Manipulation Checks. To check the primed self-con-

strual, we use the Kuhn and McPartland (1954) statementtest where participants complete 10 statements beginningwith I am. Two independent research assistants coded eachstatement as either independent or interdependent (93%agreement with any disagreements resolved through dis-cussion). Independent items include a personal description,attitude, or belief (e.g., I am intelligent). Interdependentitems refer to either a demographic group or category towhich the participant belongs (e.g., I am a Catholic) or arelationship or sensitivity to others (e.g., I am a sister). Itemsthat did not relate to either of these two categories (e.g., Iam almost done with this survey) were classified as otherand excluded from the analysis. The self-construal statementtest indicated that participants in the independent prime con-dition relative to the interdependent prime condition wrotemore individualistic sentences ( , ;M p 5.20 M p 4.50ind inter

, ), whereas those in the indepen-F(1,320) p 3.92 p !.05dent prime condition relative to the interdependent primecondition wrote fewer collectivistic sentences (M pind

, ; , ), indicating3.10 M p 5.30 F(1,320) p 8.41 p !.01interthat self-construal was successfully primed.

Brand Attitudes. The predictions were tested usingANOVA, including main effects of information valence,self-concept connection (included as a continuous variable),self-construal, brand country of origin, and all possible two-and three-way interactions of information valence, self-con-cept connection, self-construal, and brand country of origin.

The four-way interaction was also included but is not sig-nificant. Importantly, the three-way interaction of self-con-cept connection, self-construal, and information valence issignificant ( , ), which supports hy-F(1, 305) p 3.9 p !.05pothesis 1a. In addition, the three-way interaction of self-construal, brand country of origin, and information valenceis significant ( , ), supporting hypoth-F(1,305) p 4.5 p !.05esis 1b.

To explore the three-way interaction of self-concept con-nection, self-construal, and information valence further, wecreated high- and low-self-concept connection groups basedon a median split (Med p 2.20 for the foreign brand and2.30 for the local brand). Tests for simple interactions sug-gest that when an independent self-construal is primed, the

two-way interaction between self-concept connection andinformation valence is significant ( ,F(1,305) p 14.4 p !

). A simple effects test indicates no significant differences.01in mean brand evaluation in the negative information con-dition (vs. no information condition) for consumers havinghigher self-concept connection, whereas the difference issignificant for consumers having lower self-concept con-nection ( vs. 3.86; , NS;M p 3.77 F(1,305) p 0.51high

vs. 3.50; , ).M p 2.05 F(1,305) p 5.94 p !.01lowWhen an interdependent self-construal is primed, the


7/27/2019 518539

7/13


TABLE 1

STUDY 1: MEANS AND STANDARD DEVIATIONS

Independent Interdependent

No information Negative information No information Negative information

Samsung brand evaluations:Low connectedness 3.21

(.99)2.03(.48)

3.23(.75)

2.70(.85)

High connectedness 3.78(.72)

3.45(1.07)

3.98(.87)

3.11(.88)

Dell brand evaluations:Low connectedness 3.79

(.97)2.10

(1.01)3.87(.64)

3.70(.94)

High connectedness 3.96(.58)

3.90(.97)

3.95(.60)

3.85(.70)

NOTE.Cell size ranges from to .np18 np22

two-way interaction between self-concept connection andinformation valence is not significant ( ,F(1,305) p 0.74NS). Based on the median split of self-concept, meanbrand evaluation for the negative information (vs. no in-formation condition) is not significant for both higher andlower levels of self-concept connection ( vs.M p 3.20low3.60; , ; vs. 3.90;F(1, 305) p 3.00 p 1.05 M p 3.60high

, ). This pattern of results supportsF(1,305) p 3.05 p 1.05hypothesis 1a. The means and standard deviations for eachof the experimental conditions are presented in table 1.

Exploring the three-way interaction of information va-lence, brand country of origin, and self-construal further,tests for simple interactions in the primed interdependentself-construal condition indicate that the two-way interactionbetween brand country of origin and information valence issignificant ( , ). In this case, thereF(1,305) p 3.78 p !.05are no significant differences in brand evaluations for thenegative information (vs. no information context) for thelocal brand, although these differences are significant forthe foreign brand ( vs. 3.90;M p 3.79 F(1,305) plocal

, NS; vs. 3.59; ,0.28 M p 2.91 F(1,305) p 5.30 p !foreign). This pattern of results supports hypothesis 1b..01

However, when an independent self-construal is primed,the two-way interaction between brand country of originand information valence is not significant (F(1,305) p

, NS). A simple effects test in the independent local-2.10origin condition reveals significant differences in brand eval-uations for the negative information versus no informationcontext ( vs. 3.87; , ).Mp 2.99 F(1,305) p 4.80 p !.01The independent foreign-origin case also yields significantdifferences in brand evaluations for the negative infor-mation versus no information context ( vs. 3.50;Mp 2.75

, ).F(1,305) p 4.01 p !.05

Process Measures. To understand the processes un-derlying the results, we examine open-ended cognitive re-sponses from the participants. Prior research (Ahluwalia,Unnava, and Burnkrant 2001) has suggested that people whohave positive attitudes toward a target are likely to generatemore counterarguments in response to negative information.

This suggests that counterarguments are likely to mediatethe effect of negative information on brand attitudes. Further,one may also distinguish between types of counterargu-ments. Counterarguments are likely to be based on self-concept connection as well as brand country of origin. Aself-concept counterargument is defined as a counterargu-

ment presented in conjunction with a self-concept-basedthought. A brand country-of-origin counterargument is de-fined as a counterargument presented in conjunction with acountry-of-origin-based thought.

To examine the mediating role of counterarguments, twoindependent coders rated the responses as self-concept coun-terarguments, brand country-of-origin counterarguments,support arguments, and other arguments. For example, acomment such as: I am a bit skeptical of these ratings.Dells products are very good and very affordable for some-one like me was coded as a self-concept related counter-argument. An example of a brand country-of-origin coun-terargument is Dell is a reliable American brand, and I

dont trust these ratings. The coders were in agreementapproximately 95% of the time and disagreements were re-solved by discussion.

To examine the role of counterarguments, for each con-dition (independent and interdependent self-construal), thetwo variables (self-concept connection counterargumentsand brand country-of-origin counterarguments) were testedas potential mediators of the impact of self-concept con-nection and brand country of origin on brand attitude. Weperformed the mediation analysis by estimating the threeregressions recommended by Baron and Kenny (1986), withresults presented in table 2. Process measures were alsoanalyzed using ANOVA. These results indicate findings thatare consistent with the results of the mediation analysis and

are not reported for the sake of brevity.Self-concept counterarguments emerged as a mediator of

brand attitude when an independent self-construal is primed(see table 2), with this mediation supported by Sobels test( ; ). In contrast, brand countr y-of-originZp 4.11 p !.01counterarguments mediated brand attitude for participants


7/27/2019 518539

8/13


TABLE 2

STUDY 1: MEDIATIONAL ANALYSIS

Dependent variable Independent variablesStandard

beta t-value

Independent self-construal:1. Brand attitude Self-concept connection .58 5.80***

Brand country of origin (foreign p 1; local p 0) .04 .412. Self-concept-based counterarguments Self-concept connection .60 6.49***

Brand country of origin .06 .623. Brand attitude Self-concept connection .24 1.56

Brand country of origin .89 .83Self-concept based counterarguments .53 5.27**Country-based counterarguments .18 1.90

Interdependent self-construal:1. Brand attitude Self-concept connection .11 1.16

Brand country of origin (foreign p 1; local p 0) .50 5.11**2. Country-of-origin-based counterarguments Self-concept connection .05 .94

Brand country of origin .87 16.50***3. Brand attitude Self-concept connection .09 .95

Brand country of origin .35 1.20Self-concept-based counterarguments .00 .02Country-based counterarguments .24 1.99*

* .p! .05** .p! .01*** .p! .001

when an interdependent self-construal is primed (see table2), with this mediation supported by Sobels test (Zp

; ). As expected, self-concept counterarguments1.98 p !.05did not emerge as a significant mediator of brand attitudefor participants when an interdependent self-construal isprimed and brand country-of-origin counterarguments didnot emerge as a significant mediator of brand attitude forparticipants when an independent self-construal is primed.These mediation results indicate that independent consumers

resist negative information based on their self-concept coun-terarguments, whereas interdependent consumers resist neg-ative information based on brand country-of-origin coun-terarguments.

Discussion. These findings further highlight the role ofself-construal in brand attitude change. Self-construal mod-erates the relationship between self-concept connection andbrand attitudes following exposure to negative information.We demonstrate that self-construal interacts with brandcountry of origin to determine the extent to which consumersresist the negative information. The cognitive responsesshedfurther light on the mechanisms underlying the processes inthe negative information conditions. The results of the self-

concept counterarguments demonstrate that a high self-concept connection condition is associated with more self-concept counterarguments. The results show that ininterdependent states brand country-of-origin thoughts min-imize attitude change only for the local brand. To examinewhether these results are specific to the particular brandnames used in this study (Dell and Samsung), we replicatethe results with a second pair of brand names (GE and Sony).The results of this replication are consistent with the resultsreported here.

One possible mechanism driving the results is that self-construal priming may impact self-concept connectionwhich, in turn, may influence brand attitudes. To examinethis possibility, a posttest ( ) was conducted to studyn p 54the impact of self-construal priming on self-concept con-nection with the brand. The posttest was a repeated measureswithin-subjects design involving priming of self-construal.Respondents self-concept connection with Samsung andDell was measured both before and after they were exposedto either an independent or an interdependent self-construalprime. The ANOVA results with change in self-concept con-nection as a dependent measure show that self-construalpriming did not influence the change in self-concept con-nection for Samsung ( , NS) or DellF(1,52) p 0.09( , NS) (reanalyzing this data using pre- andF(1,52) p 0.10post-self-concept measures as a repeated measures variableproduced identical results).

In summary, one of the key results from this study is thatthe impact of self-concept connection varies based on self-construal using Fourniers (1998) self-concept connectionscale. One limitation is that the scale used for self-conceptconnection may potentially measure group-level connections

with the brand as well. For instance, This brand is a partof me and The brand and I have a lot in common maytap into group-level connectedness. We address this potentiallimitation in study 2 by using a modified self-concept con-nection scale.

A second issue in this study is that it was conducted inthe electronics brand context. Do the results generalize toother product categories as well? To examine this further,study 2 tests the hypotheses in a different product category,athletic shoes, which have higher average levels of self-


7/27/2019 518539

9/13


concept connection. Finally, study 2 incorporates a within-subject design to measure attitudes before and after exposureto negative brand information for the same participants.While the use of control groups in study 1 avoids the pos-sibility of demand artifacts, a within-subjects design andmeasuring the extent of attitude change for a given partic-

ipant provides a more direct measure of brand dilution. Bytesting hypotheses using an alternative experimental ap-proach, we provide additional evidence to validate the find-ings from this study.

STUDY 2

The athletic shoes category was chosen as a context forthis study for the following reasons: (1) the athletic shoecategory has strong brands, particularly with regard to theundergraduate student population; (2) athletic shoes as acategory has been used in previous brand dilution research(Ahluwalia et al. 2000); and (3) athletic shoe brands consistof both local and foreign brands.

Pretest. A pretest ( ) yielded Adidas and Niken p 30as two athletic shoe brands with both high familiarity( , ; , NS) and sim-M p 1.66 M p 1.57 t(29) p 0.44Adidas Nikeilarly high initial brand attitude ( ,M p 4.32 M pAdidas Nike

; , NS). Additionally, both Adidas and4.26 t(29) p 0.38Nike exhibited similar brand breadth ( ,M p 1.90Adidas

; , NS) and similar perceived fitM p 2.01 t(29) p 0.64Nikewith an athletic shoe ( , ;M p 4.7 M p 4.6 t(29) pAdidas Nike

, NS). The self-concept connection of both brands was0.60also measured (1 p low self-concept connection; 5 p highself-concept connection) and both brands had similar levelsof average self-concept connection ( ,M p 3.00 SD pAdidas

; , ; , NS). Similar0.91 M p 3.02 SD p 0.90 t(29) p 0.52Niketo study 1, the self-concept connection scores ranged fromvery high to very low. It is important to note that the resultsof the pretest confirmed that Adidas (a German brand) andNike (an American brand) were significantly different interms of their American-ness (Adidas p 2.4; Nike p 3.2;

, ) and in terms of their foreign-nesst(29) p 2.3 p !.05(Adidas p 2.6; Nike p 1.7; , ).t(29) p 2.78 p !.05

Main Study

We use two items that distinctly tap into individual iden-tity aspects of self-brand connection. Based on the autonomyscale from Kleine et al. (1995), a modified scale for unique

self-concept connection was developed with the followingitems: This brand makes me feel unique, and This brandis a statement of how I am different. To establish corre-spondence with study 1, we also collected participants re-sponses to the same self-concept connection scale items asthose used in study 1. The old and new scale items all loadonto the same factor, and the results of study 2 do not changeif the new scale is replaced with the old scale items.

To further validate this scale and to ensure that it doesnot measure group identity aspects of consumer-brand re-

lationships, we conducted a factor analysis of the modifiedself-concept connection scale items using pretest data. Wecollected pretest data from American students ( ) onn p 40the modified self-concept connection scale items and mea-sures of group identity (e.g., This brand makes me feelconnected with other Americans, This brand reminds me

of my country, and This brand is a statement of how Iam American). The individual- and group-identity state-ments were randomized and the scale items presented toparticipants who were asked to rate Adidas and Nike brands.A factor analysis conducted on the individual- and group-identity measures revealed a two-factor structure, corre-sponding to individual- and group-identity dimensions. Fur-ther, the factor analysis revealed that the individual- andgroup-identity measures load on separate factors. The factoranalysis results validate the use of the modified self-conceptscale to tap into unique self-concept dimensions of con-sumer-brand relationships.

Procedure. A total of 150 students (50% male; 99%ages 1824) at the University of Pittsburgh participated inthe study for either cash or course credit. No students wereof foreign origin and all students indicated at least moderatefamiliarity with the brand name. They were randomly as-signed to conditions in a 2 # 2 (brand country of origin:local vs. foreign (manipulated), self-construal: interdepen-dent or independent (manipulated)) between-subjects de-sign. Self-concept connection was measured using modifiedand original scale items.

Recall that this study focuses on establishing attitudechange following exposure to negative information. Partic-ipants were provided with a cover story which asked themto participate in two unrelated studies to be administered 1

hour apart. To measure attitude change, participants werefirst asked to indicate their evaluations of Adidas and Nike(along with their evaluations of four other brands). Thisinitial evaluation was used as a pretest score to assess theextent of attitude change. We tried to minimize potentialdemand effects from premeasurement by (1) providing aplausible cover story; (2) including other brands in the pre-measurement to minimize the possibility of making Adidasor Nike salient; and (3) allowing participants to attend toother unrelated tasks for approximately 1 hour to clear short-term memory. Based on responses to the suspicion probeat the end of the survey, students did not appear to haveany suspicions regarding the relationship between the pretestand posttest.

Self-construal priming, negative brand information, brandevaluations, and the manipulation checks in this study aresimilar to those used in study 1, with the product categorybeing athletic shoes. In this study, Puma and New Balancewere used as comparison brands for the new product. Par-ticipants were provided with some background informationregarding a new athletic shoe featuring Adidas or Nike andfollowing that were exposed to negative information (froman independent rating agency) regarding the same athleticshoe product.


7/27/2019 518539

10/13


TABLE 3

STUDY 2: BRAND ATTITUDE CHANGE

Independentattitudechange

Interdependentattitudechange

Adidas:Low connectedness .65 .89High connectedness .24 .99

Nike:Low connectedness .71 .45High connectedness .21 .43

Results

Manipulation Checks. The Kuhn and McPartland (1954)statement test was used to check primed self-construal.Statements were coded as either independent or interdepen-dent (93% agreement with any disagreements resolved

through discussion). The self-construal statement test indi-cated that participants in the independent prime conditionrelative to the interdependent prime condition wrote moreindividualistic sentences ( , ;M p 5.92 M p 4.65ind inter

, ), whereas those in the indepen-F(1,148) p 5.48 p !.01dent prime condition relative to the interdependent primecondition wrote fewer collectivistic sentences (M pind

, ; , ), indicating2.15 M p 3.96 F(1, 148) p 5.90 p !.01interthat self-construal was successfully primed.

Brand Attitude Change. The predictions were testedusing ANOVA, including main effects of self-concept con-nection (included as a continuous variable), self-construal,brand country of origin, and all possible two- and three-

way interactions of self-concept connection, self-construal,and brand country of origin. Attitude change (posttest at-titude minus pretest attitude) was used as the main dependentvariable. The use of difference scores follows earlier re-search (Ahluwalia et al. 2000). However, we also re-ana-lyzed the data using a 2 # 2 # 2 # 2 ANOVA, with pre-and posttest scores as a within-subject factor. The patternof results obtained using this alternative approach is con-sistent with the results using difference scores. For ease ofexposition, we subsequently present details of the analysiswhich use difference scores as the dependent measure.

The results with attitude change as the dependent measurerevealed that the two-way interaction of self-concept con-nection and self-construal is significant ( ,F(1,142) p 3.90

), which supports study 1 findings for hypothesis 1a.p !.05The two-way interaction of self-construal and brand countryof origin is also significant ( , ),F(1, 142) p 3.86 p !.05which supports the findings in study 1 with regard to hy-pothesis 1b. The three-way interaction of self-concept con-nection, self-construal, and brand country of origin was notsignificant ( , NS). The means are presentedF(1, 142) p 0.01in table 3.

Exploring the two-way interaction of self-concept con-nection and self-construal further, we created high- and low-self-concept connection groups based on a median split( , for both local and foreign brands). Tests forMed p 3.00simple interactions suggest that in the independent condi-tions, there is a significant difference in attitude change

between the high and low self-concept connection condi-tions ( vs. 0.68; , ).Mp 0.23 F(1,142) p 3.96 p !.05However, in the interdependent case, there are no significantdifferences between the high and low self-concept connec-tion conditions ( 1 vs. 0.67; ,Mp 0.7 F(1,142) p 0.15NS). This pattern of results is consistent with hypothesis 1aand with the previous findings in study 1.

Exploring the two-way interaction of brand country oforigin and self-construal further, when the self-construal isindependent, there are no significant differences in attitude

change (posttest brand attitude minus pretest brand attitude)for the local versus foreign brand ( vs. 0.51;Mp 0.50

, NS). However, when the self-construalF(1,142) p 0.15is interdependent, there are significant differences betweenthe local and foreign brands ( vs. 0.95;Mp 0.44

, ). This pattern of results is alsoF(1,142) p 11.0 p !.01

consistent with the results in study 1 and with hypothesis1b.

To account for potential brand ownership effects as anexplanation for the above findings, we also estimated a sep-arate model where brand ownership (1 p current owner; 0p nonowner) was included as a covariate. The brand own-ership variable was only marginally significant ( ), andp !.10the inclusion of the covariate in the model did not changethe results.

Process Measures. We examine open-ended cognitiveresponses from the participants similar to the method used instudy 1. The results of the mediation analysis are consistentwith those reported in study 1, and the details are presented

in table 4. Self-concept counterarguments emerged as a me-diator of brand attitude change for independent self-con-strual conditions, which is further supported by a Sobelstest ( ; ). Brand countr y-of-origin counter-Zp 2.36 p !.01arguments mediated brand attitude for participants in theinterdependent self-construal conditions, which is also sup-ported by a Sobels test ( ; ). These medi-Zp 3.09 p !.01ation results indicate that independent consumers resist neg-ative information based on their self-concept counterarguments, whereas interdependent consumers resist nega-tive information based on brand country-of-origin counter-arguments.

GENERAL DISCUSSION

This research illustrates how consumer-brand relationshipdimensions can affect brand equity in the face of negativeinformation, based on an individuals construal of self. Thefindings across two studies suggest that when self-conceptconnection is high, consumers tend to discount and coun-terargue the negative information, but this effect is greaterin independent self-construal conditions. Further, it appearsthat brand country-of-origin connection can significantlypromote tolerance in the face of negative information. How-


7/27/2019 518539

11/13


TABLE 4

STUDY 2: MEDIATIONAL ANALYSIS

Dependent variable Independent variableStandard

beta t-value

Independent self-construal:1. Brand attitude change (posttest minus pretest) Self-concept connection .24 2.10*

Brand country of origin (foreign p 1; local p 0) .03 .312. Self-concept-based counterarguments Self-concept connection .44 4.24***

Brand country of origin .00 .043. Brand attitude change Self-concept connection .12 .99

Brand country of origin .05 .48Self-concept-based counterarguments .30 1.98*Country-based counterarguments .10 .93

Interdependent self-construal:1. Brand attitude change Self-concept connection .04 .36

Brand country of origin (foreign p 1; local p 0) .40 3.50**2. Country-of-origin-based counterarguments Self-concept connection .02 .27

Brand country of origin .82 11.10***3. Brand attitude change Self-concept connection .09 .80

Brand country of origin .00 .02Self-concept-based counterarguments .16 1.32Country-based counterarguments .51 2.84**

* .p! .05** .p! .01*** .p! .001

ever, the role of country of origin appears to be greaterwithin interdependent self-construal conditions.

Taken together, these findings suggest a new perspectiveregarding the importance of consumer-brand relationshipsin promoting tolerance in the face of negative brand infor-mation. While past research shows that consumer-brand re-lationships can minimize the impact of negative brand in-formation on brand equity (Ahluwalia et al. 2000) or alterthe impact of brand transgressions, this research suggests

that consumer-brand relationships have multiple dimensions,and the impact of these dimensions is significantly mod-erated by self-construal. Therefore, different from previousresearch, we suggest a more nuanced perspective of howpatterns of brand dilution can vary based on consumer-brandrelationships.

Our results also shed light on the mechanisms that resultin dilution of brand equity. Despite the prior work on brandequity dilution that has focused on situations in which neg-ative information has greater (or lesser) impact on the parentbrand (Gurhan-Canli and Maheswaran 1998; John, Loken,and Joiner 1998; Swaminathan, Fox, and Reddy 2001), littlework examines how and when brand equity dilution takesplace based on these views of self-brand connectedness. The

literature on negative publicity suggests that committed con-sumers can be biased information processors (Ahluwalia etal. 2000). We provide a deeper understanding of the sourcesof bias which results in different types of counterargumen-tation (e.g., self-concept-based counterarguments vs. brandcountry-of-origin counterarguments). In summary, by de-veloping a theory-based framework of when and how self-concept connection or brand country-of-origin connectioncan influence brand equity (based on the type of self-con-strual), this research extends our current understanding of

consumer-brand relationships and their significant impact onmaintaining brand equity when challenged by negativecircumstances.

The extant research has been silent with regard to theinteraction between the dynamic self and different aspectsof consumer-brand relationships. The present research il-luminates the role of the self in the context of consumer-brand relationships by demonstrating that self-construal hasa significant impact on which aspects of consumer-brand

relationships are more important. Therefore, our researchbridges a gap in the literature by examining consumer-brandrelationships in the context of the dynamic self.

Taken together, the results indicate that brands are highlysymbolic entities that are intricately woven into the fabricof consumers lives and help shape and communicate theirindividual as well as their group identities. This dichotomousview of self-identity at the individual and group level is oneof the key assumptions of identity theory (Tajfel and Turner1979) but has not been applied to the context of consumer-brand relationships. However, the importance of ones iden-tity to consumption has been recognized (Deighton 2005)and the idea that material possessions in general act as fa-cilitating artifacts for consumer identities has been sug-

gested previously (Kleine, Kleine, and Kernan 1993, 229).For instance, Kleine et al. (1995) use attachment theory andsuggest that material possessions can be used to representdifferent facets of identity (e.g., autonomy and affiliation).Consistent with this notion, work by Escalas and Bettman(2005) suggests that brands can communicate referencegroup identity. Research by Reed (2004) highlights the roleof social identities in product judgments and attitudes. How-ever, different from past findings on this topic, the currentresearch suggests that brands help strengthen consumer iden-


7/27/2019 518539

12/13


tity at both the individual and group level. Additionally, thekey contribution of this research is specifying and experi-mentally testing the role of individual-identity (self-conceptconnection) and group-identity (brand country-of-originconnection) dimensions of consumer-brand relationshipsand the moderating role of self-construal.

This research also has various practical implications.Much of the managerial emphasis is on the role of brandsin conveying individual identities, but our research uncoversan additional basis for connecting with consumers (i.e.,brand country-of-origin connection). Other group-level as-sociations such as membership in art museums (Bhatta-charya, Rao, and Glynn 1995) and brand communities mayprovide additional bases for building relationships with con-sumers and enhancing customer retention. The impact ofnegative information on brand equity can be severe anddevastating as recent brand crises (e.g., Vioxx, Ford Ex-plorer) have indicated. The findings lead us to suggest com-munication strategies following a negative incident can po-tentially direct consumers attention to brand country of

origin to minimize brand dilution.This research is not without its limitations. One potential

limitation is that self-concept connection is a measured var-iable. In an ideal experimental design, both self-conceptconnection and self-construal should be manipulated factors.In the context of consumer-brand relationships, it is ratherchallenging to construct relationships for hypothetical brandnames in an experimental setting. Future research can focuson manipulating both self-concept connection and self-con-strual to extend and validate the results in this research.

Several interesting extensions for future research can alsobe envisaged. One possible avenue for future research maybe to examine differences in the control group receivingpositive/neutral information about a product launch rather

than no information at all. Our results regard the impact ofnegative information, but it is interesting to speculatewhether these results would hold for negative consumerexperiences, since product experience can be a valuablesource of consumer beliefs (Hoch and Deighton 1989). Ex-tending the findings from this research to other types ofproduct categories (e.g., hedonic goods) may provide ad-ditional insights. Future research should also extend ourframework to incorporate multiple roles that consumer-brand relationships play in communicating and shaping var-ious types of consumer identities (e.g., ethnic group identityor regional identity).

REFERENCES

Aaker, Jennifer L., Susan Fournier, and S. Adam Brasel (2004),When Good Brands Do Bad, Journal of Consumer Re-search, 31 (June), 116.

Agrawal, Nidhi and Durairaj Maheswaran (2005), The Effects ofSelf-Construal and Commitment on Persuasion, Journal ofConsumer Research, 31 (March), 84149.

Ahluwalia, Rohini, Robert E. Burnkrant, and H. Rao Unnava(2000), Consumer Response to Negative Publicity: The

Moderating Role of Commitment,Journal of Marketing Re-search, 37 (May), 20314.

Ahluwalia, Rohini, H. Rao Unnava, and Robert E. Burnkrant(2001), The Moderating Role of Commitment on the Spill-over Effect of Marketing Communications, Journal of Mar-keting Research, 38 (November), 45870.

Baron, Reuben and David Kenny (1986), The Moderator-Medi-ator Variable Distinction in Social Psychological Research:Conceptual, Strategic and Statistical Considerations,Journalof Personality and Social Psychology, 51 (December),117382.

Belk, Russell W. (1988), Possessions and the Extended Self,Journal of Consumer Research, 15 (September), 13968.

Bhattacharya, C. B., Hayagreeva Rao, and Mary Ann Glynn(1995), Understanding the Bond of Identification: An In-vestigation of Its Correlates among Art Museum Members,

Journal of Marketing, 59 (October), 4657.Chavis, Larry and Phillip Leslie (2005), Consumer Boycotts: The

Impact of the Iraq War on French Wine Sales in the U.S.,working paper, Stanford University, Stanford, CA, 94305.

Deighton, John (2005), Consumer Identity Motives in the Infor-mation Age, in Inside Consumption: Consumer Motives,

Goals, and Desires,ed. S. Ratneshwar and David Glen Mick,New York: Routledge, 23351.

Escalas, Jennifer Edson and James R. Bettman (2005), Self-Con-strual, Reference Groups, and Brand Meaning, Journal ofConsumer Research, 32 (December), 37889.

Fournier, Susan (1994), A Consumer-Brand Relationship Frame-work for Strategic Brand Management, unpublished disser-tation, Graduate School, University of Florida, Gainesville,FL 32611.

(1998), Consumers and Their Brands: Developing Re-lationship Theory in Consumer Research, Journal of Con-sumer Research, 24 (March), 34373.

Gurhan-Canli, Zeynep and Durairaj Maheswaran (1998), The Ef-fects of Extensions on Brand Name Dilution and Enhance-ment,Journal of Marketing Research, 35 (November), 464

73. (2000), Cultural Variations in Country of Origin Effects,

Journal of Marketing Research, 37 (August), 30917.Hoch, Stephen J. and John Deighton (1989), Managing What

Consumers Learn from Experience, Journal of Marketing,53 (April), 120.

Hong, Sung-Tai and Robert S. Wyer (1990), Determinants ofProduct Evaluation: Effects of the Time Interval betweenKnowledge of a Products Country-of-Origin and Its SpecificAttributes,Journal of Consumer Research, 17 (December),27788.

Johansson, Johnny (1989), Determinants and Effects of the Useof Made In Labels, International Marketing Reviews, 6(February), 4759.

John, Andrew and Jill Klein (2003), The Boycott Puzzle: Con-

sumer Motivations for Purchase Sacrifice,Management Sci-ence, 49 (September), 11961209.

John, Deborah Roedder, Barbara Loken, and Christopher Joiner(1998), The Negative Impact of Extensions: Can FlagshipProducts Be Diluted? Journal of Marketing, 62 (January),1932.

Klein, Jill Gabrielle, Richard Ettenson, and Marlene D. Morris(1998), The Animosity Model of Foreign Product Purchase:An Empirical Test in the Peoples Republic of China,Journalof Marketing, 62 (January), 89100.

Kleine, Robert E., III, Susan Schultz Kleine, and Jerome B. Kernan


7/27/2019 518539

13/13


(1993), Mundane Consumption and the Self: A Social-Iden-tity Perspective, Journal of Consumer Psychology, 2 (3),20935.

Kleine, Susan Schultz, Robert E. Kleine III, and Chris T. Allen(1995), How Is a Possession Me or Not Me? Character-izing Types and an Antecedent of Material Possession At-

tachment, Journal of Consumer Research, 22 (December),32743.Kuhn, Manford H. and Thomas S. McPartland (1954), An Em-

pirical Investigation of Self-Attitudes, American Sociologi-cal Review, 19 (February), 5876.

Maheswaran, Durairaj (1994), Country of Origin as a Stereotype:Effects of Consumer Expertise and Attribute Strength onProduct Evaluations, Journal of Consumer Research, 21(September), 35465.

Markus, Hazel and Shinobu Kitayama (1991), Culture and theSelf: Implications for Cognition, Emotion, and Motivation,Psychological Review, 98 (April), 22453.

Ng, Sharon and Michael Houston (2006), Exemplars or Beliefs?The Impact of Self-View on the Nature and Relative Influenceof Brand Associations, Journal of Consumer Research, 32

(March), 51929.Pomerantz, Eva M., Shelly Chaiken, and Rosalind S. Tordesillas

(1995), Attitude Strength and Resistance Processes,Journalof Personality and Social Psychology, 69 (September),40819.

Reed, Americus, II (2004), Activating the Self-Importance ofConsumer Selves: Exploring Identity Salience Effects onJudgments,Journal of Consumer Research,31 (September),28695.

Shimp, Terence A. and Subhash Sharma (1987), Consumer Eth-

nocentrism: Construction and Validation of the CETSCALE,Journal of Marketing Research, 24 (August), 28089.

Singelis, Theodore (1994), The Measurement of Independent andInterdependent Self-Construals,Personality and Social Psy-chology Bulletin, 20 (May), 58091.

Sirgy, Joseph (1982), Self-Concept in Consumer Behavior: ACritical Review,Journal of Consumer Research,9 (Decem-ber), 287300.

Swaminathan, Vanitha, Richard Fox, and Srinivas Reddy (2001),The Impact of Brand Extension Introduction on Choice,

Journal of Marketing, 65 (October), 115.Tajfel, Henri and John C. Turner (1979), An Integrative Theory

of Intergroup Conflict, in The Social Psychology of Inter-group Relations,ed. William G. Austin and Stephen Worchel,Montergy, CA: Brooks/Cole.

Trafimow, David, Harry Triandis, and Sharon Goto (1991), SomeTests of the Distinction between the Private Self and the Col-lective Self, Journal of Personality and Social Psychology,60 (May), 64955.

Triandis, Harry C. (1995),Individualism and Collectivism, Boulder,CO: Westview.

Wallendorf, Melanie and Eric Arnould (1988), My Favorite

Things: A Cross-cultural Inquiry into Object Attachment,Possessiveness, and Social Linkage, Journal of Consumer

Research, 14 (March), 53147.Weiss, Robert (1974), The Provisions of Social Relationships,

in Doing unto Others: Joining, Molding, Conforming, Help-ing, Loving,ed. Zick Rubin, Englewood Cliffs, NJ: Prentice-Hall, 1726.

Wright, Paul (1974), The Delineation and Measurement of SomeKey Variables in the Study of Friendship, Representative

Research in Social Psychology, 5, 9396.

518539

Documents