Consumers' implicit theories about personality influence their brand personality judgments

Available online at www.sciencedirect.com

Journal of Consumer Psychology 22 (2012) 545 – 557

Research Article

Consumers' implicit theories about personality influence their brand personality judgments Pragya Mathur a,⁎, Shailendra P. Jain b , Durairaj Maheswaran a

c

Zicklin School of Business, Baruch College, City University of New York, USA b Foster School of Business, University of Washington, USA c Stern School of Business, New York University, USA

Received 8 April 2009; received in revised form 6 January 2012; accepted 30 January 2012 Available online 4 February 2012

Abstract Three studies document that consumers' implicit theories about the fixedness/malleability of personality guide brand personality updating in a brand extension context. The first two studies show that extension fit with the parent brand impacts brand personality updating only for incremental (vs. entity theorists). Specifically, for incremental theorists, brand personality is enhanced (vs. diluted) when extension fit is poor (vs. good), and only when brand personality is salient. The third study identifies conditions under which entity theorists focus on brand personality. Interestingly, overall evaluations of the parent brand and extension vary only with extension fit. Implications of our research are discussed. © 2012 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved. Keywords: Implicit theories; Brand personality; Brand extensions; Information processing

Introduction Brand personality refers to the “set of human characteristics associated with a brand” (Aaker, 1997). For example, consumers associate the Cheerios brand with personality dimensions such as sincerity and competence but not with other dimensions such as sophistication and ruggedness. The kinds of personality dimensions attributed to a brand are important because they can influence the evaluation and consumption of products (Swaminathan, Page, & Gurhan-Canli, 2007). These downstream effects warrant an understanding of when consumers' perceptions of a brand's personality change. Whereas changing consumers' perceptions of a brand's personality might seem fairly simple and easy to accomplish, this may not always be the case, as some consumers may be more willing to update their personality impressions while others may be more resistant. Specifically, those who have the implicit ⁎ Corresponding author at: Zicklin School of Business, Baruch College, City University of New York, Box 12-240, One Bernard Baruch Way, New York, NY 10010, USA. Fax: +1 646 312 3271. E-mail address: [email protected] (P. Mathur).

theory that human personality is malleable (incremental theorists) seem more likely to change their beliefs about brand personality compared to those who have the implicit theory that human personality is fixed (entity theorists). We show that the differences in consumers' implicit personality theories affect brand personality judgments within the important context of brand extensions. Previous research suggests that consumers' implicit theories about the malleability of personality influence their acceptance of brand extensions (Yorkston, Nunes, & Matta, 2010). Our research builds on this line of research by investigating the effect of consumers' implicit personality theories when the parent brand and the new extension category fit well and when they fit poorly. For instance, extending the Cheerio brand to granola is a good fit because it is similar to cereal, but extending the Cheerio brand to frozen dinners is a poor fit because it is dissimilar. Rather than considering whether brand personality influences perceived fit (as do Yorkston et al., 2010), we investigate the effects of the fit between a parent brand and an extension (e.g., Cheerio's fit with frozen dinners) on changes in the parent brand's perceived personality (e.g., on Cheerio's perceived sincerity). Our research shows that fit influences the parent brand's personality beliefs of incremental

1057-7408/$ -see front matter © 2012 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jcps.2012.01.005

546

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

theorists but not entity theorists. Further, we identify the underlying mechanism. Incremental theorists are more process driven than entity theorists, so perceptions of effort expended by the parent brand to develop and introduce the extension mediate the effect on their brand personality judgments. Importantly, we investigate the effects of an extension to a new category on consumers' overall evaluation of the parent brand (e.g., whether extending Cheerios to granola enhances the overall positive evaluations of Cheerios more than extending Cheerios to frozen dinners). One might suppose that further strengthening of a brand's personality would enhance consumers' overall evaluation of the parent brand (e.g., that consumers who believe Cheerios is now even more sincere would evaluate Cheerios as a brand even more positively). However, new information does not always influence evaluations of the parent brand and brand personality inferences the same way (Johar, Sengupta, & Aaker, 2005). We show that these can be independent judgments. These differences reveal important limits on the effects of consumers' implicit theories on overall evaluations. Specifically, our research shows that only the extension's perceived fit influences the brand's overall evaluation, and not consumers' implicit theories. In sum, our research provides insights into how consumers varying in their implicit theories of personality change their perceptions of a parent brand's personality as well as its evaluations when exposed to extensions varying in fit with the parent brand.

similar to each other (extensions with a good fit) and when they are dissimilar to each other (extensions with a poor fit; Park, Milberg, & Lawson, 1991). Brand personality Brand personality results from the endowment of human characteristics and traits to brands (Aaker, 1997) allowing consumers to have relationships with brands similar to those observed in social contexts (Aaker, Fournier, & Brasel, 2004), and has been posited to be of relevance in the evaluation and consumption of brands (Swaminathan et al., 2007). Brand personality may develop based on consumers' direct and indirect interactions with the brand, as well as from the marketing efforts of firms (e.g., Fitzsimons, Chartrand, & Fitzsimons, 2008; Johar et al., 2005). In this research, we suggest that when the parent brand launches a brand extension, the extent to which parent brand personality is revised will be impacted by consumers' implicit theory of personality malleability. Specifically, we suggest that in processing brand extension information, incremental theorists will rely on their belief in the malleability of personality and update parent brand personality impressions, whereas entity theorists will focus on their belief in the fixedness of personality and will not infer personality implications of brand extensions. Brand extension fit and effort perceptions

The effect of consumer implicit theory on brand personality Implicit theories of personality malleability A significant body of social psychology research has found that individuals maintain systematically different ‘implicit theories’ about the world around them as evidenced by either a belief in the fixedness or immutability of personality, characterized by an entity theory orientation, or a belief in the changeability or malleability of personality, characterized by an incremental theory orientation (Dweck & Leggett, 1988). Implicit theories guide social interaction and information processing, and individuals rely on these theories to interpret and predict a variety of phenomena. These theories affect how individuals judge themselves, others, and objects (Elliott & Dweck, 1988). In particular, entity (vs. incremental) theorists have been observed to draw more personality inferences from behavior and focus more on performance goals, leading to a focus on outcome. Incremental theorists in contrast attribute behavior to situational factors and focus more on learning goals, emphasizing effort and mediating processes (Molden & Dweck, 2006). Recent consumer research has shown that implicit theories of human personality are also used to interpret information about marketing activity (Jain, Mathur, & Maheswaran, 2009; Park & John, 2011; Yorkston et al., 2010). For instance, consumers' implicit theories influence their acceptance of brand extensions, i.e., when brands launch new products under the existing brand name (Yorkston et al., 2010). Our research builds on these findings by examining the effect of consumers' implicit theories when the parent brand and the new extension category are

Past research has shown that consumers relate extension fit and brand effort related to developing and introducing the extension. Extensions with good fit are considered “easy” and involve “low effort” for brands due to the ease of transferability of parent brand values and attributes. In contrast, extensions with poor fit are considered “difficult” and involve “high effort” due to the increased risk of establishing the brand in a new and distant category (Ahluwalia & Gurhan-Canli, 2000). Similarly, Aaker and Keller (1990) have identified perceived difficulty of “designing or making the extension product” as a factor determining extension evaluations. The belief that extensions with poor (vs. good) fit relate to high (vs. low) effort will have differential implications for the way incremental and entity theorists judge parent brand personality. Past research has shown that while incremental theorists focus on process, entity theorists focus on the outcome (Levy, Stroessner, & Dweck, 1998). Because they focus on the high effort process, incremental theorists will appreciate extensions with poor fit. Therefore, for incremental theorists, these favorable effort perceptions associated with introducing poor fit extensions will have positive implications and will lead to an enhancement in parent brand personality impressions. In contrast, extensions with good fit will represent an easy and low effort activity for the parent brand (Ahluwalia & Gurhan-Canli, 2000) and will fail to meet the incremental theorists' effort and learning expectations, thereby leading to a dilution of parent brand personality impressions. In contrast, entity theorists, because of their focus on performance goals, which elicits an emphasis on outcomes instead of effort and process (Levy et al., 1998), will be unaffected by

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

either the effort or the personality implications of extension fit. Therefore, for entity theorists, parent brand personality impressions will be invariant. Hypothesis 1a. Incremental theorists' parent brand personality impressions will be enhanced (vs. diluted) when the brand launches an extension with poor (vs. good) fit. Hypothesis 1b. Entity theorists' parent brand personality impressions will not be influenced by extension fit. The impact of brand personality salience The accessibility of specific personality dimensions may also impact consumers' judgments of brand personality (Fitzsimons et al., 2008; Johar et al., 2005). Research has shown that brand personality is updated when the personality is salient due to its inherent association with a brand (Diamantopoulos, Smith, & Grime, 2005; Fitzsimons et al., 2008), informational cues about the brand (Johar et al., 2005; Yorkston et al., 2010, study 3), or differences in the chronic accessibility for specific traits (Fitzsimons et al., 2008; Johar et al., 2005). In general, when brand personality is salient, the personality implications of new information become apparent and people update brand personality based on the new information (Johar et al., 2005) or act in accord with the personality implications (Fitzsimons et al., 2008). When personality is not salient because the brand does not possess strong associations with one or more brand personality dimensions, or if brand information provided is not geared towards building a strong personality (Johar et al., 2005), we do not expect the individual's implicit theory to impact the parent brand personality judgments since brand personality is unlikely to be central to the judgment context. Stated formally, Hypothesis 2. Implicit theories of personality will impact the parent brand personality judgments in response to brand extensions only when brand personality is salient. The impact on extension fit on overall brand evaluations Our expectations regarding the effects of implicit theory on the overall evaluations of the parent brand and the extension follow a different pattern. New information does not always influence brand personality impressions and parent brand evaluations in the same way (Johar et al., 2005). We suggest that brand extension fit carries both personality and evaluative implications. While the personality implications of brand extension fit will impact parent brand personality inferences, the evaluative implications of extension fit will impact the overall evaluations. Past research has investigated the impact of brand extension fit on the evaluation of the extension (Loken & John, 1993; Park et al., 1991) and has found that extension fit carries evaluative implications for the extension as well as the parent brand. Specifically, extensions featuring good fit with the parent brand are more favorably evaluated than those with poor fit (Park et al., 1991). In addition, extensions with good fit can lead to more favorable overall evaluations of the parent brand as consumers consider the parent brand favorably for launching a new product

547

which is similar to the parent brand category (Gürhan-Canli & Maheswaran, 1998). In contrast, parent brand evaluations will be less favorable for extensions that have a poor fit because consumers may derive “negative inferences regarding the company's motives for introducing these poor fitting extensions and the company's manufacturing competency/expertise for these products,” (Milberg, Park, & McCarthy, 1997, p 124). Thus, the evaluative implications of the extension on the parent brand will be in line with extension fit. Therefore, for overall evaluations, we do not anticipate implicit theories to play a role since implicit theories relate only to beliefs about personality (Dweck & Leggett, 1988). We predict that the evaluative implications of extension fit is likely to impact both, the extension and parent brand evaluations in line with past findings such that extensions with a good (vs. poor) fit will be evaluated more (vs. less) favorably and will lead to more (vs. less) favorable evaluations of the parent brand (Gürhan-Canli & Maheswaran, 1998; Milberg et al., 1997). Hypothesis 3a. Extensions with good (vs. poor) fit will be evaluated more (vs. less) favorably. Hypothesis 3b. Extensions with good (vs. poor) fit will lead to more (vs. less) favorable parent brand evaluations. The above hypotheses suggest that the differential impact of implicit theories will be observed only on parent brand personality impressions (Hypothesis 1a and Hypothesis 1b) and not on overall evaluations (Hypothesis 3a and Hypothesis 3b). Implicit theory goal orientation and extension performance Past research posits that entity and incremental theorists differ not only in their beliefs about the malleability of personality, but also in their goal orientation, which impacts their consideration of success and failure (Molden & Dweck, 2006). In essence, incremental theorists, as compared to entity theorists, “(a) predominantly adopt learning goals aimed at developing and extending ability, (b) view effort as a positive thing that activates ability rather than as a negative thing that indicates a lack of ability, (c) less frequently explain failures in terms of low ability, and (d) report mastery-oriented responses of increased effort and persistence rather than helpless strategies of effort withdrawal,” (Molden & Dweck, 2006, p. 196). Since extensions with poor fit are expected to require high effort on the part of a brand (Ahluwalia & Gurhan-Canli, 2000), for incremental theorists, a successful extension with poor fit will lead to enhanced parent brand personality impressions, while an unsuccessful extension with poor fit will not impact parent brand personality adversely. For an extension with good fit, which should be relatively “easy” for a brand to succeed in (Ahluwalia & Gurhan-Canli, 2000), poor performance of the extension is likely to indicate poor effort on the parent brand's part. Thus, an unsuccessful extension with good fit will lead to parent brand personality dilution, while a successful extension with good fit is unlikely to impact personality impressions. In contrast, entity theorists focus on outcome information to judge themselves, others, and objects (Plaks, Stroessner, Dweck,

548

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

& Sherman, 2001). Consequently, they perceive positive (vs. negative) outcomes favorably (vs. unfavorably), regardless of effort or mediating processes (Dweck & Leggett, 1988) and may even become sensitive to personality in the face of extremely counterexpectant information (Plaks et al., 2001). Thus, we predict that for entity theorists, regardless of extension fit or level of effort, a successful extension will result in an enhancement of personality impressions, whereas an unsuccessful extension will result in dilution of personality impressions. Hypothesis 4a. For incremental theorists, extension success will lead to parent brand personality enhancement for extensions with poor (vs. good) fit, while unsuccessful extensions will lead to parent brand personality dilution for extensions with good (vs. poor) fit. Hypothesis 4b. For entity theorists, parent brand personality impressions will be enhanced (vs. diluted) when they encounter a successful (vs. unsuccessful) brand extension, regardless of extension fit. Cognitions We anticipate that thoughts should reflect the differential focus of incremental and entity theorists on effort and performance respectively. Incremental theorists' focus on learning goals and effort should result in more effort-related thoughts as compared to entity theorists. In contrast, entity theorists will focus on the extension performance and generate more outcome-related thoughts than incremental theorists. Hypothesis 5. Incremental (vs. entity) theorists will generate more effort-related thoughts, whereas entity (vs. incremental) theorists will generate more outcome-related thoughts. Overview Next, we report three studies which test the above predictions. The studies used a similar procedure whereby respondents indicated their personality perceptions of a real brand and learned that the brand was launching an extension varying in fit vis-àvis the parent brand (see Table 1 for parent brand and extension information details). Subsequent to extension information exposure, participants provided a series of focal measures. Findings of the three studies collectively support the above predictions. Specifically, study 1 finds the predicted interaction between implicit theories and extension fit (Hypothesis 1a and Hypothesis 1b) and incremental theorists' greater sensitivity to fit. Study 2 extends study 1 findings in two important ways. First, study 2 replicates study 1 using a different brand and product category

and thus supports the generalizability and robustness of the findings. Second, study 2 is also congenial to the expectation that parent brand personality impressions are impacted by brand extensions and consumer implicit theory only when personality is salient (Hypothesis 2). Study 3 shows that changes in the parent brand's personality, based on a successful versus a failed extension varying in fit, differ for incremental and entity theorists in accord with Hypothesis 4a and Hypothesis 4b. Studies 1 and 3 also document process findings congenial with Hypothesis 5. Finally, all three studies demonstrate that overall evaluations depend only on the extension fit and are not influenced by implicit theories (Hypothesis 3a and Hypothesis 3b). One of the key learnings of our investigation is that brand personality impressions and overall evaluations may be impacted in different ways by respondents' implicit theory, extension fit, and their interaction. Method Pretests The parent brands and extension category were selected based on pretests. In the first pretest, respondents rated several well-known brands on 15 personality traits on seven-point scales, anchored on “very low (1)/very high (7)” to generate the measure for the five brand personality dimensions (sophistication, sincerity, ruggedness, excitement, and competence; Aaker, 1997). The objective was to select a brand with strong associations with at least one or two personality dimensions. The brand Cheerios was selected as the parent brand because respondents gave it strong ratings on two personality dimensions, competence (M = 5.48), and sincerity (M = 5.59), but significantly lower ratings on the other dimensions (sophistication: M = 3.53; ruggedness: M = 2.35; and excitement: M = 3.86). In a second pretest, 43 participants examined several extension categories and indicated the degree of extension fit with Cheerios on a 7-point scale anchored on “low fit (1)/high fit (7),” with low (vs. high) fit representing extensions that had a poor (vs. good) fit with the parent brand. Granola bars and frozen dinners were chosen as the categories that had good and poor fit respectively (MGF = 5.70, MPF = 1.89, t(41) = 11.21, p b .001) respectively. A third pretest confirmed that the two extension categories did not have strong associations with any single personality dimension (all M's b 3.0). Overview and design One hundred and fifty students were part of a 2 (implicit theory: incremental vs. entity) × 2 (brand extension fit: good vs. poor) × 2 (time of measuring parent brand measures: pre-brand

Table 1 Summary of independent variables in study 1, study 2, and study 3.

Study 1 Study 2 Study 3

Implicit Theory

Brand

Extension with good fit

Extension with poor fit

Salient personality dimension

Manipulated via news article Measured via implicit theory scale Manipulated via proverbs

Cheerios Guess Timberland

Granola bar Watches Camping tents

Frozen dinner USB Sports drinks

Sincerity, competence Sophistication, competence Ruggedness

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

extension vs. post-brand extension) mixed factorial design, with timing of the measures as a within-subjects factor. Procedure The study was conducted in two parts. In the first part, participants were administered an abbreviated version of Levy et al.'s (1998) implicit theory manipulation. Participants were informed that they were taking part in a “Learning Study” and that the experimenters “wanted to understand how people learn information.” Implicit theories were manipulated by having participants read about the views of a Dr. Medin (a fictitious key note speaker at the American Psychological Association's annual conference) about the malleability or fixedness of people's personality. Respondents then answered questions about the paragraph, summarized it in their own words, and filled out the implicit theory measure of personality scale (Plaks et al., 2001). They then completed a 10-min filler task. In the second part, presented as a “Marketing Research Study,” participants read a booklet containing the brand information and extension fit manipulations. In the good (vs. poor) fit condition, respondents learned that the brand Cheerios was launching granola bars (vs. frozen dinners). Participants then completed the dependent measures, confound checks, and a suspicion probe. Dependent variables Pre and post personality measures Brand personality was measured before and after exposure to extension information, using the Aaker (1997) 15-item scale. Parent brand and extension evaluations The brand name and brand information were presented on the second page of the booklet. After reading the information, respondents evaluated the parent brand on three, seven-point scales, anchored by “very bad (1)/very good (7),” “very negative (1)/very positive (7),” and “very unfavorable (1)/very favorable (7),” which were averaged to form the pre-extension overall evaluation index (alpha = .85). Identical items were administered immediately after participants read the extension information and these measures were averaged (alpha = .87) to form the post-extension parent brand evaluation index. Respondents then evaluated the brand extension on the same scales used to evaluate the parent brand to form a brand extension evaluation index (alpha = .79). Manipulation and confound checks Respondents indicated their perceptions of extension fit on two seven-point items anchored by “is very dissimilar(1)/ similar(7) to the Cheerios brand,” and “has a low-fit (1)/has a high-fit (7).” Respondents also indicated their familiarity with the brand and extension category (“not at all(1)/very(7) familiar”), usage of the brand's products and the extension category (“very rarely(1)/very often(7)”), and certainty of attitude towards

549

parent brand and the extension (“not at all(1)/very(7) certain/ convinced/sure” respectively). Participants then indicated their involvement, motivation, and interest in the study (“not at all(1)/very(7) involved/motivated/interested,” respectively). These covariates were inconsequential to our findings and no evidence of suspicion was apparent. Additional measures As a measure of the perceived effort and difficulty of making the extension, respondents indicated their extent of agreement, anchored on “strongly disagree(1)/strongly agree(7),” with the following statements “I think that developing the [extension] requires a lot of effort by the [parent brand],” and “I think that making the [extension] is very difficult for [parent brand].” Results Manipulation and confound checks A between-subjects ANOVA on the two-item extension fit index (alpha = .92) revealed that the extension with good (vs. poor) fit was perceived to be a better fit with the Cheerios brand (MGF = 4.38, MPF = 1.98, F(1, 140) = 108.07, p b .001). To check implicit theory orientation, an individual sample t-test on the implicit theory measure scale (Plaks et al., 2001) revealed that entity and incremental theorists' scores differed significantly (alpha = .88; MIT = 4.03, MET = 2.66, t(144) = 17.57, p b .001), confirming the success of the manipulation. Parent brand personality measures The initial ratings for excitement (MIT = 2.88, MET=3.18, p=.29); competence (MIT=4.78, MET=4.89, p=.64); sophistication (MIT=3.08, MET=3.28, p=.41); sincerity (MIT=5.10, MET=4.77, p=.23); and ruggedness (MIT=2.42, MET=2.69, p=.39) dimensions were similar for the two theorists (see Table 2 for alphas). A mixed-design ANOVA on the sincerity dimension yielded a three-way interaction between extension fit, respondent implicit theory, and parent brand personality measures before and after the extension was introduced (F(1, 141) = 4.33, p b .05). Further analysis yielded a two-way interaction between the timing of the parent brand measure and fit (F(1, 141) = 4.99, p b .01) only for incremental theorists, who reported dilution of brand sincerity for the extension with good fit (MGF-Pre = 5.12, MGF-Post = 4.58, F(1, 141) = 6.93, p b .01). In contrast, the extension with poor fit led to enhanced sincerity impressions (MPF-Pre = 5.09, MPF-Post = 5.47, F(1, 141) = 3.45, p b .05). For entity theorists, there was no impact on parent brand personality impressions (MGF-Pre = 4.96, MGF-Post = 4.75; MPF-Pre = 4.62, MPF-Post = 4.27; all Fs b 1) (Table 3). An identical analysis on ‘competence’ revealed a three-way interaction of time, theory, and fit (F(1, 141) = 6.56, p b .01). An interaction between the timing of the parent brand measure and fit was observed only for incremental theorists (F(1, 141) = 3.45, p b .05) for whom the parent brand's competence impressions

550

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

Table 2 Summary of reliability of personality measures in study 1, study 2, and study 3. Dependent measures:

Study 1

Cronbach alpha of personality measures

Pre extension

Post extension

Pre extension

Study 2 Post extension

Experimental group

Control group

Sincerity Down-to-Earth Honest Wholesome Cheerful Excitement Daring Spirited Imaginative Up-to-date Competence Reliable Intelligent Successful Sophistication Upper class Charming Ruggedness Outdoorsy Tough Strong

0.73 ✓ ✓ ✓ ✓ 0.73 ✓ ✓ ✓ ✓ 0.74 ✓ ✓ ✓ 0.69 ✓ ✓ 0.73 ✓ ✓ –

0.77 ✓ ✓ ✓ ✓ 0.76 ✓ ✓ ✓ ✓ 0.73 ✓ ✓ ✓ 0.71 ✓ ✓ 0.71 ✓ ✓ –

0.81 ✓ ✓ ✓ ✓ 0.85 ✓ ✓ ✓ ✓ 0.79 ✓ ✓ ✓ 0.73 ✓ ✓ 0.79 ✓ ✓ –

0.85 ✓ ✓ ✓ ✓ 0.84 ✓ ✓ ✓ ✓ 0.80 ✓ ✓ ✓ 0.72 ✓ ✓ 0.82 ✓ ✓ –

– – – – – – – – – – – – – – 0.68 ✓ ✓ 0.80 ✓ ✓ ✓

– – – – – – – – – – – – – – 0.68 ✓ ✓ 0.81 ✓ ✓ ✓

were diluted for the extension with good fit (MGF-Pre = 5.06, MGF-Post = 4.51, F(1, 141) = 4.90, p b .05), but were marginally enhanced for the extension with poor fit (MPF-Pre = 4.17, MPF-Post = 4.49, F(1, 141)= 3.01, p = .08). For entity theorists, no changes in personality impressions were observed (MGF-Pre = 4.83, MGF-Post = 4.64; MPF-Pre = 4.81, MPF-Post = 4.52; all Fs b 1). No other effects were significant. These findings supported Hypothesis 1a and Hypothesis 1b. Similar analyses on the other brand personality dimensions (excitement, ruggedness, sophistication) did not yield any differential effects of implicit theories and brand extensions (all ps N .25). Thus, implicit theory impacted only those brand personality dimensions that were focal for the Cheerios brand. Parent brand evaluations Incremental and entity theorists did not differ on initial brand evaluations (MIT = 5.41, MET = 5.19, p = .87). A 2 (implicit theory)× 2 (extension fit) × 2 (time of measuring parent brand evaluation: pre-brand extension vs. post-brand extension) mixeddesign ANOVA on overall evaluations revealed a two-way interaction between time of measurement and fit (F(1, 141) = 29.98, p b .01) and a main effect of the time of measuring parent brand evaluation (F(1, 141)= 28.72, p b .001). Further analyses revealed that extensions with poor fit led to lower evaluations for both theorists (incremental: MPF-Pre = 5.29, MPF-Post = 4.26, F(1, 141) = 58.31, p b .001; entity: MPF-Pre = 5.14, MPF-Post = 4.46, F(1, 141) = 14.89, p b .001). Thus, regardless of theory, extensions with poor fit led to less favorable parent brand evaluations post-extension (MPF-Pre = 5.21, MPF-Post = 4.36, F(1, 141) = 70.79, p b .001). This result supports Hypothesis 3b and confirms that overall evaluations post-extension were driven by fit alone and were independent of participant implicit theory.

Study 3

Brand extension evaluation A 2 (implicit theory) × 2 (extension fit) between-subjects ANOVA on the extension evaluation index revealed a main effect of fit: consistent with Hypothesis 3a, the extension with good (vs. poor) fit was evaluated more favorably (MGF = 4.74, MPF = 3.22; F(1, 142) = 64.46, p b .001). No other effects were significant. Additional measures The extension with the poor (vs. good) fit was associated with greater perceived effort (MGF = 4.22, MPF = 5.06, F(1, 141) = 11.23, p b .01). This effect was qualified by an interaction between theory and fit (F(1, 141)= 4.75, p b .05) such that only incremental theorists associated the extension with poor (vs. good) fit with greater effort (MIT-GF = 3.96, MIT-PF = 5.33, F(1, 141) = 11.57, p b .01). We observed a similar pattern of results on perceived difficulty of making the extension. The extension with poor (vs. good) fit was perceived as more difficult to make (MGF = 3.78, MPF = 5.17, F(1, 141) = 29.86, p b .001), and this effect was qualified by an interaction theory and fit (F(1, 141) = 13.77, p b .001). Incremental theorists considered the extension with poor (vs. good) fit to be more difficult to make (MIT-GF = 3.29, MIT-PF = 5.63, F(1, 141) = 31.82, p b .001). Moderated mediation analysis To examine whether perceived effort mediated the relationship between extension fit and personality updating for incremental theorists, but not for entity theorists, we conducted a moderated mediation analysis (Berger & Fitzsimons, 2008). A linear regression found that extension fit predicted perceived

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557 Table 3 Study 1: personality measures, parent brand evaluations, extension evaluations, and other measures for Cheerios. Dependent measures Incremental theorists Sincerity impressions Pre extension Post extension Competence impressions Pre extension Post extension Parent brand evaluation Pre extension Post extension Brand extension evaluation Perceive effort Perceive difficulty Manipulation check implicit theory a Manipulation check extension fit b Cell size Entity theorists Sincerity impressions Pre extension Post extension Competence impressions Pre extension Post extension Parent brand evaluation Pre extension Post extension Brand extension evaluation Perceive effort Perceive difficulty Manipulation check implicit theory a Manipulation check extension fit b Cell size

Extension with good fit

Extension with poor fit

5.12 4.58

5.09 5.47

5.06 4.51

4.17 4.49

5.53 5.51 4.65 3.96 3.29 4.05 4.54 36

5.29 4.26 3.39 5.33 5.63 4.10 2.08 38

4.96 4.75

4.62 4.27

4.83 4.64

4.81 4.52

5.24 5.28 4.82 4.49 4.27 2.73 4.22 38

5.14 4.46 3.06 4.78 4.71 2.66 1.89 33

Note: extension with good Fit: Granola bars; extension with poor fit: frozen dinner. a Results of the implicit theory scale measure. b Results of the composite fit index.

effort (t(144) = 2.79, p b .01). A multiple regression was run to predict the updating of the sincerity dimension (measured as the difference between the sincerity personality measure after and before the extension) based on the mediator (perceived effort), the moderator (implicit theory), the independent variable (extension fit), and the interaction between the mediator and the moderator. We found that the effect of perceived effort on personality updating was impacted by implicit theory as shown by the significant mediator x moderator interaction (t(142) = 2.31, p b .05). Mediation effects of perceived effort were observed for incremental theorists (indirect effect; z = 1.38, p b .05) but not entity theorists (indirect effect; z = .27, p N .50). Further, a separate mediation analyses for incremental and entity theorists showed that for incremental theorists, extension fit significantly impacted both, perceived effort (β = 1.35, p b .001) and the sincerity impression difference (β = .53, p b .05). Also, effort impacted the sincerity impression difference (β = .29, p = .07). When we included both, effort and extension fit in a regression with the difference in sincerity as the dependent variable, we saw that the effects of perceived effort remained significant, (β = .54,

551

p b .05), but that of extension fit was not significant (β = −.02, p N .50). Therefore, perceived effort mediated the effect of extension fit on the updating of sincerity for incremental theorists. A similar analysis for entity theorists revealed that effort did not mediate the updating of sincerity. Specifically, fit did not impact perceived effort (β = .23, p = .28) or the sincerity impression difference (β = −.27, p N .70), and effort did not impact the sincerity impression difference (β = .04, p N .70). When we included both terms in a regression predicting the sincerity difference, neither extension fit (β = .24, p = .31) nor perceived effort (β = .04, p = .97) were significant predictors. We observed a similar pattern of results when we examined the effects of perceived effort on the updating of the competence dimension. Therefore, our analysis supports the notion that perceived effort mediates the relationship between extension fit and personality impressions but only for incremental theorists. Discussion Study 1 supports the expectation that implicit theories influence brand personality impressions. Specifically, only incremental theorists, who believe in the malleability of personality, modified their parent brand personality impressions in response to extension information, whereas entity theorists' personality impressions remained unchanged. In support of Hypothesis 1a and Hypothesis 1b, incremental theorists varied their personality impressions in line with their perceptions about the effort undertaken by the parent brand such that extensions with a good (vs. poor) fit led to brand personality dilution (vs. enhancement). A moderated mediation analysis showed that effort mediated the relationship between extension fit and brand personality updating for incremental, but not entity theorists. Interestingly, as outlined in Hypothesis 3a and Hypothesis 3b, both theorists' overall evaluations of the parent brand and the extension were consistent with the extension fit such that the extension with good (vs. poor) fit led to more (vs. less) favorable extension and parent brand evaluations. In sum, while parent brand personality was impacted by implicit theory and extension fit, overall evaluations were affected by extension fit alone. Study 2 extends these findings in two important ways. First, by altering the salience and accessibility of the parent brand personality dimensions, study 2 establishes that high personality salience and accessibility will lead to updating of brand personality only for incremental theorists. However, when brand personality is not salient, the observed difference between the changes in brand personality impressions and overall evaluations will be attenuated. Second, study 2 also extends and replicates the findings of study 1 with a different brand (Guess) and a different salient personality dimension (sophistication), thereby establishing the generalizability of the findings. Method Design and Procedure Two hundred and fifteen respondents participated in study 2, which was a 2 (implicit theory) × 2 (extension fit) × 2 (brand

552

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

personality salience: high versus low) between subjects design. Study 2 was similar to study 1, with two important exceptions. First, half the respondents were induced to have high salience for a specific brand personality dimension, and second, implicit theory was measured using the implicit theory of personality scale (Plaks et al., 2001). Independent variables For this study, it was important to choose a brand that did not a priori have strong associations with any single personality dimension because, when consumers perceive that a brand has relatively weak association with a personality dimension (low salience), that dimension should not impact or be impacted by other factors (Fitzsimons et al., 2008; Johar et al., 2005). Further, choosing a brand with weak personality dimension associations would allow us to manipulate stronger personality dimension salience consistent with our study objectives. Pretests using the 15-item brand personality scale resulted in the selection of the brand Guess, which was rated low on all brand dimensions (sincerity: M = 2.72; excitement: M = 2.40; competence: M = 2.30; sophistication: M = 2.35; and ruggedness: M = 1.07). The salience of the parent brand's focal personality dimension, sophistication, was manipulated based on Aaker et al. (2004) and Johar et al. (2005). In the high salience condition, the brand (Guess) was described as possessing several traits related to sophistication. In the low salience condition, Guess was described using several traits that were unrelated to a single personality dimension. After completing the relevant dependent measures, respondents indicated how much they thought about the parent brand's personality (“not at all(1)/very much(7)”), which served to verify the validity of the salience manipulation. A pretest generated the extension categories for Guess. The goal was to select categories which differed on fit perceptions, but were similar across other important variables (likeability and familiarity) and had similar personality associations that differed from the focal dimension (sophistication). This was done to ensure that any observed changes in the parent brand personality would arise only due to the interactive effect of extension fit and implicit theory and not due to any direct effects of the extension category personality on the parent brand. Based on these pretests, watches and USB drives were chosen as the categories with good (vs. poor) fit (MGF = 6.19, MPF = 1.77, t(50) = 12.74, p b .001). Importantly, these categories were perceived as being similar on all personality dimensions and had the strongest association with the competence dimension (MWatches-Competence = 5.52, MUSB-Competence = 5.53, t b 0). Results Implicit theory Implicit theory orientation was determined using the implicit theory measure scale (Plaks et al., 2001). Thirty eight respondents could not be classified and were removed from further

analysis. A t-test on the remainder revealed that the two theorists differed significantly on their implicit theory scores (MIT = 4.35, MET = 2.49, t(175) = 28.76, p b .001).

Manipulation and confound checks A between-subjects ANOVA on the two-item extension fit index (alpha = .96) revealed that the watch category (vs. USB drive) was considered to have good (vs. poor) fit with the Guess brand (MGF = 4.19, MPF = 1.65, F(1, 168) = 132.59, p b .001) and would perform better (MGF = 6.03, MPF = 4.28, F(1, 168) = 23.17, p b .001). The extension with good (vs. poor) fit was also was associated with lower (vs. higher) effort by the parent brand (MGF = 4.53, MPF = 5.15, F(1, 169) = 5.04, p b .05). Finally, respondents thought more about the parent brand's personality in the salient (vs. not salient) condition (MSalient = 5.01, MNot-Salient = 2.28, F(1, 168) = 18.68, p b .001). Covariates were not significant and will not be discussed further.

Parent brand personality measures The two focal dimensions were sophistication, which was related to the parent brand personality description, and competence, which was related to the personality associations of the extension categories. The initial parent brand sophistication and competence impressions did not differ for the two theorists (sophistication: MIT = 3.93, MET = 3.97; competence: MIT = 3.75, MET = 3.77; ps N .50). As expected, as a result of the salience manipulation, the sophistication ratings were higher in the salient (vs. non-salient) condition (MSophistication-Salient = 4.30, MSophistication-Not Salient = 3.86, F(1, 168) = 5.74, p b .05). A 2 (implicit theory) × 2 (extension fit) × 2 (brand personality salience) × 2 (time of measuring parent brand measures) mixeddesign ANOVA on the sophistication rating yielded a four-way interaction (F(1,167) = 7.12, p b .01) and a three-way interaction of time, salience, and fit (F(1,167) = 5.36, p b .05). Further analysis revealed a two-way interaction between time and fit (F(1, 167) = 9.33, p b .01) when brand personality was salient. Additional analysis revealed an interaction between time and fit (F(1, 167) = 5.93, p b .05) only for incremental theorists. Decomposing these higher-order interactions revealed that only for incremental theorists, when personality was salient, parent brand sophistication impressions were enhanced for the extension with poor fit (MPF-Pre = 4.29, MPF-Post = 5.43, F(1, 167)= 9.32, p b .01), but were diluted for the extension with good fit (MGF-Pre = 4.30, MGF-Post = 2.95, F(1,167) = 12.38, p b .01). When personality was not salient, no significant effects were observed (all Fs b 1). Thus, in support of Hypothesis 2, implicit theories of personality impacted parent brand personality in response to brand extensions only when brand personality was salient. A similar analysis on the parent brand competence impressions revealed no significant results. This could be because more and repeated data points from its extension(s) may be necessary before the parent brand's personality impressions undergo measurable and perceptible change (Table 4).

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557 Table 4 Study 2: personality measures, parent brand evaluations, extension evaluations, and other measures for Guess. Personality salient Dependent measures Incremental theorists Sophistication impressions Pre-extension Post-extension Parent brand evaluations Pre-extension Post-extension Brand extension evaluations Perceived effort Manipulation check personality salience a Manipulation check extension fit b Cell size Entity theorists Sophistication impressions Pre-extension Post-extension Parent brand evaluations Pre-extension Post-extension Brand extension evaluations Perceived effort Manipulation check personality salience a Manipulation check extension fit b Cell size

Personality not salient

Good fit

Poor fit

Good fit

Poor fit

4.30 2.95

4.29 5.43

3.89 3.95

3.91 3.87

4.53 4.48 3.85 3.84 5.06 4.18 21

4.58 3.69 3.32 5.10 5.11 1.68 25

4.35 4.30 3.84 4.45 2.43 4.17 20

4.31 3.33 2.46 5.13 2.36 1.65 21

4.29 4.38

4.35 4.38

3.71 3.67

3.91 4.00

4.41 4.48 4.21 4.88 4.89 4.22 23

4.43 3.98 3.34 5.08 4.97 1.67 23

4.49 4.42 3.59 4.95 2.14 4.21 22

4.45 3.90 2.69 5.32 2.17 1.59 22

Note: Extension with good fit: watches; extension with poor fit: USB. a Results on manipulation check on salience of Guess brand sophistication. b Results of the composite fit index.

553

Discussion Studies 1 and 2 offer evidence that implicit theories impact the updating of brand personality, and that incremental theorists' personality impressions are diluted (vs. enhanced) for extensions with good (vs. poor) fit. In addition, findings of these two studies support the suggestion that the effects of implicit theory and extension fit on brand personality impressions are evident only when brand personality is made salient, either through the strong association of a brand with a personality dimension (study 1) or through the brand description (study 2). Study 3 more deeply examines the differences between the two theorists in their process versus outcome focus (Hypothesis 5) observed in study 1. Based on this difference, it also identifies extension failure as an important condition under which entity theorists may show sensitivity to brand personality. Our expectation is that while incremental theorists' focus on process will lead them to value the perceived effort in launching an extension more than the extension's success or failure (Hypothesis 4a), entity theorists' focus on the outcome will cause them to be impacted by negative outcomes such as extension failure (Hypothesis 4b). Method Overview and design One hundred and ninety seven university students participated in a 2 (theory) × 2 (extension fit) × 2 (extension performance: successful vs. unsuccessful), between-subjects design. An additional fifty participants were part of a control condition. Procedure

Parent brand evaluations Incremental and entity theorists did not differ on initial evaluations of the parent brand (MIT = 4.21, MET = 4.32, p N .50). A 2 (theory) × 2 (extension fit) × 2 (salience of brand personality) × 2 (time of measuring parent brand measures) mixed-design ANOVA on overall evaluations revealed a two-way interaction between time of measurement and fit (F(1, 168)= 17.90, p b .001). Regardless of salience, the extension with poor fit led to less favorable parent brand evaluations for both theorists (incremental: MPF-Pre = 4.44, MPF-Post = 3.51, F(1, 168)= 14.20, p b .001; entity: MPF-Pre = 4.38, MPF-Post = 3.62, F(1, 168) = 5.03, p b .05), thereby supporting Hypothesis 3b and converging with findings in study 1.

The study was conducted in two parts. In the first part, implicit theory was primed (Poon & Koehler, 2006). In the second part, participants saw the parent brand information, the extension fit, and extension performance manipulations. Respondents then completed the Plaks et al. (2001) implicit theory scale which served as a manipulation check, and responded to a suspicion probe (no evidence of suspicion was observed). Participants in the control condition were given only the parent brand information and their overall evaluations and brand personality impressions were measured. Independent variables Implicit theory

Brand extension evaluations A 2 (theory) × 2 (extension fit) × 2 (brand personality salience) between-subjects ANOVA on the brand extension evaluation index revealed only a main effect of fit, such that the extension with good (vs. poor) fit was evaluated more favorably (MGF = 3.87, MPF = 2.95, (F(1, 169) = 22.20, p b .001). No other effects were significant.

Implicit theory was manipulated using proverbs stressing the fixedness or malleability of human traits. Respondents were told that the first part of the study related to an effort to understand how people used proverbs in their everyday lives. Respondents in the entity theory condition read the three proverbs, “You cannot teach an old dog new tricks,” “Old habits die hard,” and “A leopard cannot change its spots,” whereas those in the incremental theory condition read three different proverbs, “It is never too

554

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

late to learn,” “Experience is the best teacher,” and “When in Rome, do as Romans do.” Subsequently, participants indicated their level of familiarity with each proverb, restated the proverb in their own words, suggested three different situations in which the proverb could be applied, and indicated the name of a person who exemplified the proverb. This extensive elaboration on each proverb has been shown to orient individuals as entity or incremental theorists (Poon & Koehler, 2006). Brand extension fit Participants then read about the Timberland brand, which was carefully selected because pretests revealed that the brand had a strong association with a single personality dimension (ruggedness, M = 5.18). Participants learned that Timberland was a well-known company, “committed to delivering high quality products that can benefit people's lives everywhere,” and “dedicated to improving people's lives through a commitment to innovation, quality, value, and design, with a focus on creating comfortable, cutting edge products.” Respondents then read that Timberland had recently launched a new product (camping tents: good fit; or sports drinks: poor fit). Pretests with fifty respondents confirmed that Timberland had a good (vs. poor) fit with the tent (vs. sports drinks) category (MTents = 5.31, MSportDrinks = 1.69, t(48)= 14.32, p b .001). The categories did not differ significantly on their personality dimensions (excitement: MTent = 4.46, MSD = 4.84; competence: MTent = 4.23, MSD = 4.67; sophistication: MTent = 4.40, MSD = 4.69; sincerity: MTent = 3.79, MSD = 3.72; and ruggedness: MTent = 4.31, MSD = 4.45; all ps N .50), thereby ensuring that any observed changes in the Timberland personality in response to the extensions would not be caused by the differential personality effects of the extension categories. Extension performance Extension performance information was presented in the Consumer Reports format. Respondents saw the new Timberland product compared to three brands (Tents: Paha Que, Eureka, and Wenzel; Sports drinks: Taut, Full Throttle, and Si Hypotonic) (Ahluwalia & Gurhan-Canli, 2000). The comparisons were made on two dimensions: “ruggedness” and “sophistication.” In the successful (vs. unsuccessful) information condition, the extension received a higher (vs. lower) ruggedness rating than all other brands in the category, establishing that the extension was successful (vs. not successful). The sophistication ratings remained the same and comparable across conditions.

Parent brand evaluations All respondents (including control) indicated their evaluation of Timberland on six seven-point scales, anchored on “bad (1)/ good (7),” “low quality (1)/high quality (7),” “not likeable (1)/ very likeable (7),” “unpleasant (1)/pleasant (7),” “unappealing (1)/appealing (7),” and “inferior product (1)/superior product (7),” which were averaged to form a brand evaluation index (alpha = .94). Extension fit was checked as in study 1 (alpha = .89). Information valence was checked by respondents indicating whether the information provided featured “few positive attributes (1)/many positive attributes (7),” “few negative attributes (1)/many negative attributes (7),” (reverse scored) and “inferior to competing brands (1)/superior to competing brands (7)” (alpha = .81). Additionally, as in studies 1 and 2, respondents indicated the perceived effort associated with extension. Cognitions Participants then listed all thoughts that came to mind while reading the new product information. Thoughts were coded by an independent coder (unaware of the hypotheses) as outcome thoughts (O) if they related to the success or failure of the brand extension (e.g., “These camping tents are better than the competition”); effort thoughts (E) if they related to the effort of launching the extension (e.g., “Soft drinks must be difficult to launch for Timberland”), and irrelevant thoughts (e.g., “Timberland is available in many stores”). Results Manipulation checks A 2 (implicit theory) × 2 (extension fit) × 2 (extension performance information) between-subjects ANOVA on the performance valence index revealed only a main effect of performance information, such that extension success information conveyed more positive and fewer negative features, and portrayed the Timberland branded product as superior to other products (MSuccessful = 4.55, MUnsuccessful = 3.22, F(1,197) = 47.15, p b .001). An ANOVA on the fit measures revealed that the extension with good (vs. poor) fit was evaluated as being more similar to and typical of Timberland (alpha = .93, MGF = 4.05, MPF = 3.24, F(1, 199) = 18.16, p b .001). The implicit theory check revealed that theorists differed on their implicit theory scale scores (alpha = .85; MIT = 4.47, MET = 2.40, t(198) = 22.50, p b .001). Parent brand personality measures—ruggedness

Dependent variables Parent brand personality measures The target dimension of “ruggedness” was estimated by averaging the measures for tough, strong, and outdoorsy traits (alpha = .82; Aaker, 1997). Participants in the control group also completed the same measures. See Table 2.

The hypotheses were tested by contrasting the ruggedness impressions of the experimental group with the control group (Ahluwalia & Gurhan-Canli, 2000). Using the error term from the omnibus ANOVA, planned contrasts with the control group revealed that for incremental theorists, parent brand ruggedness impressions were enhanced for the successful extension with poor fit (MControl = 5.69; MIT-PF-Success = 6.31, t(198) = 2.91,

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

p b .05), but the successful extension with good fit had no impact on parent brand personality (MIT-GF-Success = 5.67, t b 1). Similarly, incremental theorists' brand ruggedness impressions was diluted when they saw information about the unsuccessful extension with good fit (MIT-GF-Unsuccessful = 4.09, t (197) = −4.96, p b .001), but not when they encountered the same information for the extension with poor fit (MIT-PF-Unsuccessful = 5.63, t b 1). In accord with Hypothesis 4a, incremental theorists' ruggedness impressions were enhanced (vs. diluted) for the successful (vs. unsuccessful) extension with poor (vs. good) fit. Further, entity theorists' parent brand ruggedness impressions were diluted when they learned that both the good and poor fit extensions were unsuccessful (MControl-Ruggedness = 5.69; MET-GF-Unsuccessful = 4.65, t(198) =−3.88, p b .001; MET-PF-Unsuccessful = 4.58, t(198) = −3.54, p b .01), but not when the extensions were successful (MET-GF-Success = 5.75; MET-PF-Success = 5.77, all ps N .50). Thus, our results partially supported Hypothesis 4b(Table 5).

555

that the extension with poor (vs. good) fit was associated with greater perceived effort (MGF = 4.20, MPF = 5.01, F(1, 192) = 12.71, p b .01) and incremental theorists perceived greater overall effort (MIT = 4.89, MET = 4.32, F(1, 192) = 6.24, p b .05). These effects were qualified by a two-way interaction between extension fit and performance (F(1, 192) = 4.52, p b .05). No other effects were significant. Mediation analysis To examine whether perceived effort mediated the relationship between extension fit, extension performance, and the change in the ruggedness impressions for incremental theorists, we conducted a moderated mediation analysis as in study 1. As expected, while perceived effort mediated the relationship between extension fit and performance for incremental theorists, no such mediation was observed for entity theorists. Cognitions

Parent brand evaluations An ANOVA on thoughts revealed the two expected main effects of implicit theory—incremental (vs. entity) theorists generated more effort related thoughts (MIT-Effort = 0.49, MET-Effort = 0.31; (F(1, 189)= 4.37, p b .05), while entity (vs. incremental) theorists generated more outcome related thoughts (MET-Outcome = 1.76, MIT-Outcome = 0.73; F(1, 189) = 7.23, p b .01). Thus, in support of Hypothesis 5, incremental (vs. entity) theorists focused on the extension effort (vs. outcome).

Planned contrasts with the control group revealed that for incremental theorists, regardless of fit, parent brand evaluations declined in response to the unsuccessful extension (MIT-GF-Unsuccessful =4.49, t (1,198) = − 2.01, p b .05; MIT-PF-Unsuccessful = 4.34, t(198) = − 2.56, p b .05), a finding replicated for entity theorists (MControl-Eval = 5.18; MET-GF-Unsuccessful = 4.41, t (198) = −2.69, p b .01; MET-PF-Unsuccessful = 4.31, t(198) = −2.76, p b .01). Parent brand evaluations did not improve in any conditions (all ts b 1).

Discussion Additional measures In study 3, consistent with studies 1 and 2, implicit theories affected parent brand impressions only for relevant personality dimensions, in this case, ruggedness. Further, while incremental

A 2 (theory) × 2 (fit) × 2 (extension performance) between subjects ANOVA on the perceived effort measure showed

Table 5 Study 3: personality impressions, parent brand evaluations, and other measures for Timberland.

Dependent measures by theorists Incremental theorists Ruggedness dimension measure Parent brand evaluations Perceived effort Manipulation check for implicit theory a Manipulation check for extension performance b Manipulation check for extension fit c Cell sizes Entity theorists Ruggedness dimension measure Parent brand evaluations Perceived effort Manipulation check for implicit theory a Manipulation check for extension performance b Manipulation check for extension fit c Cell sizes

Extension with good fit

Extension with poor fit

Successful

Unsuccessful

Successful

Unsuccessful

Control condition

5.67 5.45 4.83 4.34 4.67 4.04 23

4.09 4.49 3.96 4.61 3.04 4.02 25

6.31 5.30 5.19 4.47 4.59 3.20 26

5.63 4.34 5.58 4.48 3.30 3.16 24

5.69 5.18

5.75 5.34 4.16 2.23 4.58 4.13 24

4.65 4.41 4.84 2.44 3.42 4.01 25

5.77 5.20 4.45 2.56 4.35 3.26 24

4.58 4.31 4.81 2.38 3.11 3.34 26

5.69 5.18

Note: extension with good fit: camping tents; extension with poor fit: sports drinks. a Results of the implicit theory scale measure. b Results of the extension performance valence index. c Results of the composite fit index.

50

556

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

theorists' parent brand personality impressions were affected by their perceptions of effort expended by the parent brand, entity theorists' impressions were impacted only when the extension had failed. Thus, study 3 extended studies 1 and 2 and provides novel insight into entity theorists' outcome focus and suggests conditions that may elicit the entity theorist's sensitivity to personality. While we predicted that the outcome focus of entity theorists would result in updating of all brand related information for both positive and negative outcomes, we found that they weighted negative outcomes (unsuccessful extension) more than positive outcomes (successful extension). Both, cognitive responses and mediation analysis supported greater process (effort) focus for incremental (vs. entity) theorists and greater outcome focus for entity (vs. incremental) theorists. Finally, the effects of implicit theories on a different focal dimension, “ruggedness,” confirmed the robustness of our findings. General discussion This research shows that parent brand personality is influenced by consumers' implicit theories of personality in the context of brand extensions. We show that brand extension fit with the parent brand can have both, personality and evaluative implications for the parent brand. Further, this extension fit information interacts with a consumer's implicit theory, resulting in systematic influences on brand personality impressions. Specifically, personality implications of extension fit influence incremental theorists, but not entity theorists. Incremental theorists are more sensitive to personality information, focus on the process (effort), and view extensions with poor (vs. good) fit more favorably in updating parent brand impressions. Entity theorists are outcome focused, consider evaluative (vs. personality) implications of extension fit, and update their personality impressions only in response to an outcome, especially failure. Three studies provided convergent evidence for implicit theory effects and the underlying mechanisms across different operationalizations of implicit theory, personality salience, and personality dimensions. Our research adds to the implicit theory literature by identifying a condition under which entity theorists will be sensitive to brand personality—when the outcome is negative. In such situations, entity theorists' evaluative as well as brand personality impressions may both be diluted. Also, the difference between changes in overall evaluations and personality impressions may create an inconsistency for incremental theorists. There is some support for this asymmetry in other domains where it has been observed that incremental, but not entity theorists, may be more comfortable with responding to and holding inconsistent information (e.g. Plaks et al., 2001, p 883). Our results show that at a given point in time, this may indeed be the case. Our research also contributes to brand personality research by demonstrating that brand personality may be impacted by brand extensions differently from what was previously thought. Specifically, we show that extensions have significant personality implications beyond their well-documented evaluative impact. Our work extends emerging evidence of the impact of implicit theory on brand extensions (e.g. Yorkston et al., 2010) by identifying perceived effort associated with introducing an extension as an

important factor that impacts the effects of brand extensions on brand personality, especially for one segment of consumers—incremental theorists. Our specific contribution to brand extension research is the demonstration that while extensions with good fit with the parent brand may have a favorable evaluative impact, they are at risk of diluting parent brand personality impressions. This research also has implications for the effect of implicit theory on brand personality beyond the context of brand extensions. For instance, our findings suggest that incremental theorists, being more sensitive to personality information, may be a focal target for marketing activity aimed at highlighting changes in brand personality. In contrast, entity theorists may be better focal targets for marketing activity highlighting the steadfast nature of brand personality. Our work may extend to new product launches in general in terms of differences in the adoption of innovation for incremental and entity theorists: incremental theorists may adopt new products relatively quickly if they perceive greater effort associated with them, whereas entity theorists may adopt a wait-and-watch strategy, regardless of the effort. This research also identifies implicit theory as a moderator of the effects of the new information on brand personality ratings and overall evaluations, suggesting that the observed asymmetry between personality impressions and overall evaluations exists, but only for incremental theorists. This observed asymmetry raises three interesting questions. First, what are the conditions when this asymmetry may not obtain? Brands and products featured in our studies had relevant and important product features as well as personalities. However, offerings may have strong personality associations with less salient features (e.g., high end luxury products) or highly salient features and weak or difficult-toarticulate personality associations (e.g., light bulbs). In such cases, the observed asymmetry between brand personality impressions and overall evaluations may not emerge. Second, what may be the medium and long term consequences of a company launching a series of extensions with poor fit over multiple periods? One possibility is that if a brand were to launch successful extensions with poor fit (Park et al., 1991), the brand may be perceived as more extendible (e.g. Gürhan-Canli, 2003), and perceptions of fit may be altered. For example, General Electric (GE) operates successful extensions in technologybased categories such as lighting, aircrafts, etc. When GE introduced financial services, this would have been considered a poor fit. Subsequently, when GE extended into real estate, this category no longer represented a poor fit for GE because of its association with financial services. Therefore, what appeared to be a poor fit or disparate category for GE with its earlier extensions may now be considered more suitable and a good fit for the company. Such situations may be better explained by a more dynamic model of the interactive effects of implicit theory and fit on parent brand personality impressions. Along the same lines, while incrementalists may hold opposing views relating to brand personality and overall evaluations and at a given point in time, they may resolve these views in a multi-period scenario as perceptions of fit, and hence perceived effort, change. It is possible that perceived effort may impact both, brand personality and overall evaluations in such settings when growing success in categories with poor fit is attributed to high effort by incrementalists. Over time, the

P. Mathur et al. / Journal of Consumer Psychology 22 (2012) 545–557

combination of extension success and effort perceptions may align incrementalists' personality impressions and evaluations. Finally, the impact of extension category personality associations may also affect the parent brand personality over time. As the parent brand's portfolio of extensions with disparate fit changes over time, the strength of the parent brand's personality association may weaken, thereby making the parent brand more likely to be impacted by extension category personality associations. Third, it is possible that the observed difference between brand personality and overall evaluations may arise because the overall parent brand evaluation may reflect a judgment about the brand's characteristics, while incremental theorists' personality impressions may reflect human personality inferences related to the organization's capability perceptions and the managers' goals. While this implication has merit, extant research suggests that while organization capabilities influence perceptions of effort, these perceptions directly impact brand perceptions (Aaker & Keller, 1990; Ahluwalia & GurhanCanli, 2000). Therefore, it appears that perceptions of the brand and organization are correlated in this domain, and the organization and brand are not independent. However, future research may examine this asymmetry under conditions when the brand and organization are perceived as independent entities, such as multi-brand conglomerates or brand alliances. Finally, it can be argued that perceptions of effort should impact specifically the sincerity dimension and not seemingly unrelated dimensions such as sophistication, since sincerity is often determined by the effort expended. Our results are consistent with past findings that have documented that only the most accessible and focal personality dimensions get updated (study 1: sincerity and competence; study 2: sophistication; study 3: ruggedness). Nevertheless, perceived effort may also have semantic effects on brand personality. For instance, Aaker et al. (2004) found that brand transgressions impact consumer-brand relationships differently for brands with strong sincerity versus strong excitement dimensions. Similarly, in our conceptualization, it is possible that perceived effort could have impacted several personality traits directly related to effort. However, if these personality traits were not germane and salient to the brand, it is likely that these impressions may not have been updated. Future research can examine the semantic effects of mediating variables such as perceived effort and perceptions of difficulty of developing and launching extensions on specific personality dimensions when all traits are equally salient. References Aaker, J. (1997). Dimensions of brand personality. Journal of Marketing Research, 34(3), 347–356. Aaker, J., Fournier, S., & Brasel, S. A. (2004). When good brands do bad. Journal of Consumer Research, 31(1), 1–16.

557

Aaker, D., & Keller, K. L. (1990). Consumer evaluations of brand extensions. Journal of Marketing, 54(1), 27–41. Ahluwalia, R., & Gurhan-Canli, Z. (2000). The effects of extensions on the family brand name: An accessibility–diagnosticity perspective. Journal of Consumer Research, 27(3), 371–381. Berger, J., & Fitzsimons, G. (2008). Dogs on the street, pumas on your feet: How cues in the environment influence product evaluation and choice. Journal of Marketing Research, 45(1), 1–14. Diamantopoulos, A., Smith, G., & Grime, I. (2005). The impact of brand extensions on brand personality: Experimental evidence. European Journal of Marketing, 39(1–2), 129–149. Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to motivation and personality. Psychological Review, 95(2), 256–273. Elliott, E. S., & Dweck, C. S. (1988). Goals: An approach to motivation and achievement. Journal of Personality and Social Psychology, 54(1), 5–12. Fitzsimons, G. M., Chartrand, T. L., & Fitzsimons, G. J. (2008). Automatic effects of brand exposure on motivated behavior: How Apple makes you “think different”. Journal of Consumer Research, 35(3), 21–35. Gürhan-Canli, Z. (2003). The effect of expected variability of product quality and attribute uniqueness on family brand evaluations. Journal of Consumer Research, 30(1), 105–114. Gürhan-Canli, Z., & Maheswaran, D. (1998). The effects of extensions on brand name dilution and enhancement. Journal of Marketing Research, 35(4), 464–473. Jain, S. P., Mathur, P., & Maheswaran, D. (2009). The influence of consumers' lay theories on approach/avoidance motivation. Journal of Marketing Research, 46(1), 56–65. Johar, G. V., Sengupta, J., & Aaker, J. (2005). Two roads to updating brand personality impressions: Trait versus evaluative inferencing. Journal of Marketing Research, 42(4), 458–469. Levy, S. R., Stroessner, S. J., & Dweck, C. S. (1998). Stereotype formation and endorsement: The role of implicit theories. Journal of Personality and Social Psychology, 74(6), 1421–1436. Loken, B., & John, D. R. (1993). Diluting brand beliefs: When do brand extensions have a negative impact? Journal of Marketing, 57(3), 71–84. Milberg, S. J., Park, C. W., & McCarthy, M. S. (1997). Managing negative feedback effects associated with brand extensions: The impact of alternative branding strategies. Journal of Consumer Psychology, 6(2), 119–140. Molden, D. C., & Dweck, C. S. (2006). Finding “meaning” in psychology: A lay theories approach to self-regulation, social perception, and social development. American Psychology, 61(3), 192–203. Park, J.K. & John, D.R. (2011). Capitalizing on brand personalities in advertising: The influence of implicit self-theories on ad appeal effectiveness. Journal of Consumer Psychology. Park, C. W., Milberg, S. J., & Lawson, R. (1991). Evaluation of brand extensions: The role of product level similarity and brand concept consistency. Journal of Consumer Research, 18(2), 185–193. Plaks, J. E., Stroessner, S. J., Dweck, C. S., & Sherman, J. W. (2001). Person theories and attention allocation: Preferences for stereotypic versus counterstereotypic information. Journal of Personality and Social Psychology, 80(6), 876–893. Poon, C. S. K., & Koehler, D. J. (2006). Lay personality knowledge and dispositionist thinking: A knowledge-activation framework. Journal of Experimental Social Psychology, 42(2), 177–191. Swaminathan, V., Page, K., & Gurhan-Canli, Z. (2007). “My” brand or “Our” brand: The effects of brand relationship dimensions and self-construal on brand evaluations. Journal of Consumer Research, 34(2), 248–259. Yorkston, E., Nunes, J., & Matta, S. M. (2010). The malleable brand: The role of implicit theories in evaluating brand extensions. Journal of Marketing, 74(1), 80–93.