Competing for Consumer Identity:Limits to Self-Expression and thePerils of Lifestyle Branding

66 Journal of Marketing Vol. 75 (May 2011), 66–82 © 2011, American Marketing Association ISSN: 0022-2429 (print), 1547-7185 (electronic) Alexander Chernev, Ryan Hamilton, & David Gal Competing for Consumer Identity: Limits to Self-Expression and the Perils of Lifestyle Branding The idea that consumers use brands to express their identities has led many companies to reposition their products from focusing on functional attributes to focusing on how they fit into a consumer’s lifestyle. This repositioning is welcomed by managers who believe that by positioning their brands as means for self-expression, they are less likely to go head-to-head with their direct competitors. However, the authors argue that by doing so, these companies expose themselves to much broader, cross-category competition for a share of a consumer’s identity. Thus, they propose that consumers’ need for self-expression through brands is finite and can be satiated when consumers are exposed to self-expressive brands. Moreover, they argue that consumers’ need for self-expression can be satiated not only by a brand’s direct competitors but also by brands from unrelated product categories, nonbrand means of self-expression, and self-expressive behavioral acts. The authors examine these propositions in a series of five empirical studies that provide converging evidence in support of the notion that the need for selfexpression can be satiated, thus weakening preferences for lifestyle brands. Keywords: brand, preferences, identity, self-expression, lifestyle, competition Alexander Chernev is Associate Professor of Marketing (e-mail: ach@ northwestern.edu), and David Gal (e-mail: d-gal@northwestern.edu) is Assistant Professor of Marketing, Kellogg School of Management, Northwestern University. Ryan Hamilton is Assistant Professor of Marketing, Goizueta Business School, Emory University (e-mail: ryan_hamilton@ bus.emory.edu).The authors thank Andrea Bonezzi, Aaron Brough, Pierre Chandon, Karsten Hansen, Vincent Nijs, and the participants at the 2008 INSEAD marketing camp for their valuable comments. This research was supported by MSI Research Grant 4-1507. One of the significant marketing changes in the past decade involves the dramatic increase in the variety of ways in which consumers can express their identities. A key driver of this change has been the growth of one-to-one marketing and mass customization, which has not only expanded accessibility to traditional means of selfexpression (e.g., hobbies, cheering for favorite sport teams and music groups, wearing iconic brands) but also introduced a variety of innovative self-expressive formats. A recent vehicle for self-expression involves the ability to customize company offerings to reflect individuals’ personal preferences. For example, Coke recently introduced its Freestyle fountain drink machines, providing customers with the option to mix selections from more than 100 choices into a custom beverage, and Starbucks allows customers to create their own beverage from more than 50,000 possible combinations. In addition to expressing their identity through choice, customers often engage in acts of selfexpression by customizing products to reflect their identity. Such self-expressive acts are ubiquitous. They can involve selecting the ringtone of a mobile phone or customizing the background of a computer or mobile phone. Consumers can customize the face of a credit card, and the look of a T-shirt, sneakers, or a backpack. People can even customize the content of a classic novel by embedding themselves into the plot. The rapid growth of social media and peer-to-peer communications present another opportunity for self-expression. Facebook, YouTube, and Twitter provide customers with an environment in which they can voice their opinions and find other people that share the same interests, thus enriching their social identity. Virtual reality games, such as Second Life and The Sims, not only allow customers to express their identity but also enable them to create new personas. Recent developments in Internet technology and mobile communications have further contributed to customers’ ability to express their identity by allowing customized media content 24 hours a day, seven days a week. Portable multimedia devices such as the iPod, iPhone, and iPad— along with the likes of Pandora Radio, Rhapsody, and Napster—have made it easy for consumers to listen to their music of choice, follow the performance of their favorite sports team, and watch their favorite television shows whenever and wherever they choose. This unprecedented access to personally relevant self-expressive offerings and information raises the question of how this increase in the means of self-expression—both in terms of number and access—sways consumer preferences. In particular, this research examines how the availability of alternative means of self-expression influences consumer preferences for selfexpressive brands. Prior research has documented the role of brands as a means of self-expression in various contexts, such as the role of brands as identity signals (Berger and Heath 2007), the role of reference groups as a source of brand meaning (Escalas and Bettman 2005), the nature of the consumer– brand relationship (Aggarwal 2004; Fournier 1998; Kleine, Kleine, and Kernan 1993), the role of brands as indicators of prestige value (Braun and Wicklund 1989), and the strength of consumers’ emotional attachments to brands (Thomson, MacInnis, and Park 2005). However, despite this extensive research, questions involving the extent to which brands can serve as a means of self-expression and the boundaries of expressing one’s identity through brands have remained largely unexplored. In fact, most prior research has implicitly assumed that consumers’ need to express their identities is constant and does not depend on the number of available means to fulfill this need [e.g., Amaldoss and Jain 2005; Braun and Wicklund 1989; Kim and Sherman 2007). To illustrate this point, consider several consumers with identical preferences: One consumer recently purchased a Nautica jacket, a second one just came back from a concert by his favorite band, a third was browsing the branded apparel in a shopping mall, a fourth just finished watching a baseball game featuring her favorite team, and a fifth was playing with LEGOs, building an action figure of Iron Man, his favorite action hero. Now imagine these consumers contemplating the purchase of an unrelated branded item (e.g., a Movado watch). Would their brand preferences be affected by their previous actions? In particular, would they believe the Movado brand to be as personally relevant as a consumer who did not engage in any of these activities? Would their prior behavior affect the degree to which they differentiated the Movado brand from other watch brands? Would their willingness to pay be influenced by their prior activities? Conventional wisdom suggests that consumers’ brand preferences are not likely to be affected by their actions in unrelated product categories and/or domains. In the context of the preceding example, this implies that as long as consumers have identical preferences, the perceived brand uniqueness, the relevance of the Movado brand to consumers, and their willingness to pay for this brand should be the same. In contrast, we show that this is not the case and that consumer brand preferences are a function of the activities they were involved in prior to evaluating a given brand—more specifically, the degree to which these prior activities afforded the opportunity to express their identities. This argument is based on the notion that consumers’ need for self-expression is finite and ultimately can be satiated, such that the value consumers place on self-expressive brands tends to decrease as the number of alternative means of self-expression increases. Furthermore, we propose that consumers’ reliance on a given brand to express their identity not only depends on brands in the same category but is also a function of the availability of alternative means of self-expression, such as brands in unrelated product categories and even nonbrand means of self-expression, including relevant affiliations, hobbies, and social interactions. To illustrate, a consumer’s preference for the Ralph Lauren brand is likely to be a function of the strength of the consumer’s associations with crosscategory self-expressive brands (e.g., Apple, Swatch, Whole Competing for Consumer Identity / 67 Foods Market) or even whether the consumer has recently engaged in self-expressive activities such as creating a work of art, updating a Facebook profile, and donating to a philanthropic organization. We outline the theory leading to these predictions in more detail in the following sections. Theoretical Background Brands as Means of Self-Expression Brands are commonly defined as marketing tools created for the purpose of differentiating a company’s offering from the competition and creating value for target customers (Keller 2007). Brands create value for customers on two dimensions: by serving to signal the quality of the underlying offerings (Wernerfelt 1988) and creating meaningful associations that add value beyond the intrinsic product attributes (Fournier 1998; Gardner and Levy 1955). The increased degree of product commoditization in the past two decades, stemming from standardization of technological design and manufacturing processes, has made brand associations—in particular, associations related to one’s self-identity—an increasingly important source of brand value. A central tenet of this research is that consumers use brands to express and validate their identity (Aaker 1997; Berger and Heath 2007; Escalas and Bettman 2005). This proposition stems from the more general notion that individual behavior is motivated by the need to reaffirm selfimage (Dunning 2005; Rogers 1947). In this context, prior research has shown that brands are often valued to the extent that they reaffirm people’s principles or beliefs (Kleine, Kleine, and Kernan 1993; Levy 1959; Solomon 1983). The self-expressive function of brands can be related to the notion of conspicuous consumption, a term used to describe the acquisition of products mainly for the purpose of attaining or maintaining social status (Veblen 1899). Typically, conspicuous consumption involves lavish spending on brands for the purpose of self-expression by displaying income or wealth. The notion that consumers conspicuously use brands that display their knowledge of culture, taste, or style has received further support from recent consumer research (Amaldoss and Jain 2005; Twitchell 2002). Brands can also be used to communicate membership in particular social or professional groups, through both the use of brands that signal membership in desirable groups (Braun and Wicklund 1989; Escalas and Bettman 2005; Wicklund and Gollwitzer 1981) and the avoidance of brands that signal membership in undesirable groups (Berger and Heath 2007). Furthermore, brands have been shown to convey otherwise hidden aspects of a consumer’s self-image because consumers frequently choose brands that they consider appropriate for the image they have of themselves (Dolich 1969; Tucker and Painter 1961). In addition to serving as an external signal, brands can serve to establish and confirm a consumer’s self-concept and identity without explicitly aiming to attain social status, recognition, or acceptance (Belk 1988; Fournier 1998). In this case, people’s motivation to express their inner states is guided by the desire to signal their self-identity not to others but to themselves (Bodner and Prelec 2003; Loewenstein 1999). This self-signaling aims to reaffirm people’s perception of the type of person they are. The idea of selfsignaling as a driver of people’s decisions is consistent with the notion that by revealing their preferences, consumers derive self-diagnostic utility from choice, allowing them to discover or define their own preferences (Akerlof and Kranton 2000). In this context, it has been shown that consumers tend to prefer brands that are more similar to their ideal self-concept in the context of conspicuous consumption but prefer brands more similar to their actual self-concept when the consumption is inconspicuous (Ross 1971). Identity Saturation and Brand Preferences The finding that people use brands to express their identities raises the question of identifying factors that influence a brand’s ability to serve as a means of self-expression and, in particular, the boundaries of the self-expressive capacity of brands. Most prior research has implicitly assumed that a brand’s ability to serve as a means of self-expression is primarily a function of brand-specific factors (e.g., a brand’s symbolic value) and that it does not change with the availability of other means of self-expression. In contrast, we argue that a consumer’s need for self-expression can be temporarily satiated whenever consumers express themselves. As a result, there are limits to their use of brands to express their identities. This means that as consumers express their identities—through brands or other means— their need for further self-expression tends to decline along with their preferences for subsequently evaluated selfexpressive brands. To illustrate, a consumer who associates strongly with a brand because it fulfills the need for selfexpression might have weaker preferences for other selfexpressive brands relative to someone who has not established a close relationship with a self-expressive brand. Moreover, merely making consumers aware of the different means they already use to express their identity is likely to decrease their need to self-express in subsequent choices. Thus, we argue that when a consumer’s need for selfexpression has already been met by some brands, additional self-expressive brands are relatively less attractive. The proposition that a consumer’s need for self-expression is finite builds on the more general principle that consumers seek to fulfill their needs and may experience satiation when those needs have been met. Previous research has shown that when a need is strong, people value the means of satisfying that need; when the need is weak, the same means are valued less (Lewin 1935). For example, consumers are willing to pay more for food when hungry (Wertenbroch and Skiera 2002), smokers value cigarettes more when deprived of nicotine (Brendl, Markman, and Messner 2003), and students standing in line to pay their tuition are willing to pay more for a tuition-waiver lottery than for a lottery with an equivalent cash payout (Markman and Brendl 2000). Satiation has been invoked to explain behavior in a variety of domains. For example, research on consumer variety seeking (Kahn and Wansink 2004; Lattin and McAlister 1985; Simonson 1990) has indicated that as consumers 68 / Journal of Marketing, May 2011 become satiated on the configuration of attributes specific to one offering, the value they derive from repeated consumption of the preferred offering decreases to the point where other alternatives become more attractive. Prior research has documented satiation effects in a variety of domains including food (Inman 2001), massages (Nelson and Meyvis 2008), music (Ratner, Kahn, and Kahneman 1999), television (Nelson, Meyvis, and Galak 2009), art (Berlyne 1971), and consumer products (McAlister 1982; Redden 2008). Recognizing that some of the value consumers derive from brands comes from meeting a more general need for self-expression implies that consumer brand preferences are a function of the availability not only of other self-expressive brands but also of all alternative means of expressing identity. In this context, we propose that the extent to which consumers use brands to express their identities is not limited to self-expressive brands in the same category but is also a function of the availability of alternative means of expressing identity. Such means can include self-expressive brands in unrelated product categories, nonbrand selfexpressive items, and self-expressive behavioral acts. For example, the introduction of Apple’s iPhone may have weakened the personal relevance and brand preferences not only of other smart phone brands but also of brands in unrelated categories, such as apparel, food, and fashion accessories. Likewise, to the degree that articulating preferences for nonbrand items (e.g., a favorite book, pastime, vacation spot) expresses a consumer’s identity, it is also likely to weaken a consumer’s brand preferences. We examine the impact of satiation on consumers’ need for self-expression on three key dimensions of brand preferences: a brand’s personal relevance, a brand’s perceived uniqueness, and consumers’ willingness to pay for a particular brand. A brand’s personal relevance captures the self-expressive function of a brand by focusing on the degree to which consumers perceive a brand to be related to their identity and to which they have closer personal relationships than with other brands (Aaker 1997; Aaker, Fournier, and Brasel 2004; Fournier 1998). Stronger preferences for a self-expressive brand are likely to be reflected in greater perceived personal relevance of the brand and a closer perceived relationship with the self-expressive brand. A brand’s perceived uniqueness reflects the degree to which consumers view brands to be differentiated from one another (Amaldoss and Jain 2005; Berger and Ward 2010). Thus, stronger preferences for a self-expressive brand should be associated with greater perceptions of brand differentiation. Finally, willingness to pay represents the behavioral outcome associated with the strength of a consumer’s brand preferences (Thaler 1985; Wertenbroch and Skiera 2002). The stronger a consumer’s preference for a brand, the more a consumer should be willing to pay for that brand. To summarize, we argue that a person’s need to express his or her identity is finite and tends to be satiated through alternative means of self-expression. As a result, the value consumers place on self-expressive brands is likely to decrease as the number of alternative means of self-expression increases. We further posit that consumer preferences for a particular self-expressive brand can be weakened due to satiation not only by brands in the same product category but also by brands in unrelated categories, by nonbrand items, and by self-expressive behaviors. We expect identity saturation to influence brand preferences in three key aspects: decrease a brand’s personal relevance, increase a brand’s perceived similarity to other brands, and decrease a brand’s monetary valuation (willingness to pay) by consumers. We test the preceding predictions in a series of five experiments that examine the impact of the availability of alternative means of self-expression on consumer brand preferences and investigate some of the key factors that moderate this effect. In the first experiment, we examine whether merely asking consumers to articulate their favorite brands weakens their subsequently elicited brand preferences in unrelated product categories. Building on the findings from this study, the second experiment investigates whether nonbrand means of self-expression (e.g., identifying favorite sports teams, songs, hobbies) are likely to decrease the personal relevance of subsequently evaluated brands. This experiment also shows that the effects of identity saturation are more pronounced for symbolic than functional brands. Experiment 3 demonstrates that merely evaluating a series of self-expressive brands is likely to decrease consumer preferences for unrelated brands. Experiment 4 further documents that the proposed identity saturation effect is a function of the strength of a person’s need for self-expression, such that it is more pronounced in the presence of a threat to self-identity. Finally, Experiment 5 demonstrates that engaging in a self-expressive behavior, such as product customization, can satiate a person’s need for self-affirmation, subsequently weakening preferences for self-expressive brands. We use converging measures of consumer brand preferences—brand relevance, brand similarity, and willingness to pay—across all five experiments to document the identity saturation effect and explore its underlying process. More specifically, our first experiment examines whether brand preferences can be influenced simply by asking consumers to think about unrelated brands that they are already using to express their identities. If, as we hypothesize, articulating one’s favorite brands is likely to lower the need for self-expression, then merely thinking about relevant self-expressive brands will lower the value the consumer places on a brand’s ability to serve as a means of self-expression. Therefore, Experiment 1 tests the proposition that a person’s brand preferences are a function of the self-expressive capacity of previously adopted brands, such that increasing the salience of brands already used to express one’s self- identity will decrease preferences for subsequently evaluated brands. Experiment 1 The primary goal of Experiment 1 is to test the proposition that a person’s reliance on brands as a means of selfexpression is likely to decrease as the awareness of alternative but unrelated means of self-expression increases. In particular, this experiment examines whether asking people to express their identity by articulating their favorite brands Competing for Consumer Identity / 69 tends to weaken their preferences for subsequently evaluated brands. Method Respondents were 102 students randomly assigned to one of two scenarios. Respondents assigned to the first scenario were asked to think of brands that were personally very important to them and to list up to eight brands they considered the most relevant (high self-expression condition). In contrast, respondents assigned to the second scenario were asked to think of brands that were personally relevant to their parents and to list up to eight brands their parents considered the most relevant (low self-expression condition). We chose parents as a group from whom respondents were likely to dissociate themselves, such that identifying brands relevant to this group was likely to strengthen respondents’ need for self-expression in the subsequent brand evaluation task. Following the self-image articulation task, respondents evaluated a set of brands in five product categories. Each evaluation task consisted of five brands: backpacks (The North Face, JanSport, Columbia, Mountain Hardwear, Timbuk2), watches (Movado, Swatch, Rolex, Seiko, Omega), deodorant (Gillette, Degree, Sure, Right Guard, Ban), sunglasses (Oakley, Foster Grant, Revo, Giorgio Armani, RayBan), and jeans (Levi’s, Lucky Brand, Lee, Calvin Klein, Versace). We measured strength of preferences for the available brands by asking respondents to evaluate brands in terms of their personal relevance (Broniarczyk and Alba 1994). In particular, respondents allocated 100 points among the five brands based on the degree to which they perceived them to be personally relevant. At one extreme, respondents who perceived all brands to be equally relevant would allocate an equal amount of points (20) to each brand. At the other extreme, respondents who perceived a particular brand to be very personally relevant would allocate 100 points to their most preferred brand and 0 points to the other brands. Results A key premise of the experimental manipulation is that articulating personally relevant brands is likely to be associated with a greater degree of self-expression and, thus, more likely to decrease respondents’ need for subsequent selfexpression than articulating outgroup-relevant brands. To check the validity of this assumption, we administered the self-expression task to a separate group of 39 respondents from the same population and asked them to generate a list of up to eight brands that were either very personally relevant to themselves or very personally relevant to their parents. We tested the impact of the brand elicitation task on respondents’ need for self-expression by asking them to draw a stick figure picture of themselves both before and after the brand elicitation task. We conjectured that if respondents’ need for self-expression had been (at least partially) satiated by the brand-elicitation task, their drawings of themselves following the brand-elicitation task would decrease in size, whereas the opposite effect would be observed if respondents’ need for self-expression had been strengthened by the brand-elicitation task. This prediction is based on prior research showing that more acute needs can lead people to place greater value on the means of fulfilling the need (Markman and Brendl 2000; Wertenbroch and Skiera 2002) and that this value can sometimes be manifested through larger physical representations of the means of satiating the need. For example, smokers deprived of nicotine judged the true length of a standard cigarette to be longer than did people who had recently smoked (Brendl, Markman, and Messner 2003), and children from poor families judged coins to be larger than did children from wealthy families (Bruner and Goodman 1947). The data show that respondents who listed self-relevant brands subsequently drew smaller representations of themselves than they did before the brand-listing task (mean height difference = –5.3 mm, range: –24 mm to 14 mm, SD = 1.0). In contrast, participants who listed brands relevant to someone else subsequently drew larger self-representations than they did before the brand-listing task (mean height difference = 4.6 mm, range: 14 mm to 76 mm, SD = 2.1). We tested the significance of these data by examining a model in which the height of respondents’ drawings was given as a function of the drawing sequence (before or after the selfexpression task) and the nature of the self-expression task (low vs. high self-expression). The data show that the difference in self-representations across the two brand-elicitation conditions was significant, as indicated by the significant interaction effect (F1, 37 = 3.54, p < .05). Moreover, the difference in the self-representations among respondents in the self-expression condition was significant (F1, 37 = 5.16, p < .05), indicating that the manipulation created a difference in the need to self-express. With regard to the main hypothesis, we argue that invoking personally relevant brands in the minds of the respondents weakens their preference for subsequently evaluated brands. We operationalized the value placed on the target brands as (1) the absolute score of the brand perceived to be most personally relevant and (2) the point difference (or “distance”) between the most relevant and second most relevant brand. Each of the 102 respondents evaluated brands in the five product categories, which yielded 490 ratings (20 missing data points): 234 in the high self-expression condition and 256 in the low self-expression condition. The data show that respondents in the high self-expression condition, who were asked to articulate personally relevant brands, perceived the subsequently presented brands as being more similar in terms of their personal relevance than did those in the low self-expression condition, who were asked to identify brands relevant to their parents. In particular, the ratings of the brand perceived to be personally most relevant were lower for respondents in the high self-expression condition than for those in the low self-expression condition (M = 45.1, SD = 23.0 vs. M = 55.2, SD = 28.2; F1, 100 = 7.96, p < .01). This effect was consistent across all five product categories, as revealed by a nonsignificant interaction between the factors of interest and product category (F4, 380 = .92). An additional measure of consumers’ perceptions of brand relevance is the perceived distance between the most relevant and second most relevant brand. Comparing con70 / Journal of Marketing, May 2011 sumer evaluations of their most preferred and second most preferred brand (in terms of personal relevance) offers a better indication of dispersion of preferences across the available brands. To illustrate, consider the following two sets of ratings: 50-50-0-0-0 and 45-15-15-15-10. If we consider only the highest rating, it seems that consumer preferences are stronger in the first scenario (50 > 45). However, the relative valuations of the brands in these sets suggest that consumer preference is likely to be stronger in the second scenario because one of the brands is preferred over the other by a higher margin (50 – 50 = 0 < 45 – 15 = 30). In this context, we interpret a greater difference between the two most preferred brands as an indication of stronger preferences. The data show that the distance between the highest rated and the second highest rated brands was smaller for respondents who were asked to articulate personally relevant brands than for respondents who were asked to articulate their parent’s preferences (M = 20.7, SD = 27.7 vs. M = 33.2, SD = 38.4; F1, 100 = 7.29, p < .01). This effect was consistent across all five product categories, as revealed by a nonsignificant interaction between the factors of interest and product category (F4, 380 = .71). This finding is consistent with the proposition that increasing the salience of personally relevant brands tends to weaken subsequent consumer brand preferences. Discussion The data this experiment furnishes lend support to the proposition that increasing the prominence of self-expressive brands that are already a part of a consumer’s identity is likely to weaken future brand preferences, as reflected in ratings of the personal relevance of these brands. We also show that this effect is not limited to brands in the same product category but also can be caused by articulating selfexpressive brands in unrelated product categories. An important aspect of the theory advanced in this research is that brands compete not only with other brands for a share of consumers’ identity but also with nonbrand means of self-expression. We argue that because self-concept operates on a more general level than brand-specific effects, consumers can express their identity through a variety of alternative means that go beyond brands. Building on the notion that people’s need for self-expression is finite, we posit that the availability of nonbrand means of self-expression can also weaken preferences for subsequently evaluated brands. Thus, we predict that a person’s brand preferences are a function of the self-expressive capacity of previously adopted nonbrand means of self-expression, such that increasing the salience of nonbrand means already used to express self-identity tends to decrease people’s preferences for subsequently evaluated brands. Prior research has argued that certain brands are more likely to be used as means of self-expression than others (Aaker 1997; Katz 1960; Richins 1994). Building on this research, we distinguish two types of brand associations: functional and symbolic. Functional associations relate the meaning of the brand to functional aspects of the underlying product or service, such as physical characteristics, performance, and reliability, whereas symbolic associations add value beyond the intrinsic product attributes. For sim- plicity, we refer to brands heavy on functional associations as functional brands and brands heavy on symbolic associations as symbolic brands. In the context of our theorizing, the distinction between symbolic and functional brands implies that the identity saturation effects observed in the first experiment are likely to be a function of the self-expressive nature of the brands evaluated. Indeed, if the decline in the perceived brand relevance reported in the first experiment can be attributed to a decline in consumers’ need for self-expression, we would expect this effect to be contingent on the degree to which brands enable consumers to express their identity. Thus, we predict that the decrease in brand preferences reported in the first experiment will be more pronounced in the context of symbolic than functional brands. This line of reasoning leads to the prediction that the impact of the availability of alternative means of self-expression on consumer preferences for brands is a function of the brands’ self-expressive capacity, such that it will be more pronounced for symbolic than utilitarian brands. We test the validity of this proposition in the following experiment. Experiment 2 Building on the findings from the first experiment, our goal in the second experiment is to investigate whether the strength of consumer brand preferences is a function of the availability of nonbrand means of self-expression. Experiment 2 also tests the strength of the identity saturation effect as a function of brands’ ability to fulfill consumers’ need for self-expression. Method Respondents, 104 participants in an executive seminar, were randomly assigned to one of two self-expression conditions. Respondents in the high self-expression condition were asked to write down their favorite sports teams, television shows, books, hobbies, and products. In contrast, respondents in the low self-expression condition were given an Competing for Consumer Identity / 71 unrelated filler task that did not ask for the delineation of any personally relevant preferences. In addition to manipulating the level of self-expression, we also varied the degree to which individual brands had the potential to fulfill people’s need for self-expression. Following the initial task, respondents in both self-expression conditions evaluated brands in four of eight different product categories: sneakers (Nike, Adidas, Puma, Converse, and Reebok), knit shirts (Polo, Nautica, Abercrombie & Fitch, and Lacoste), jackets (The North Face, Timberland, Patagonia, and Columbia), watches (Seiko, Movado, Omega, and Rolex), soft drinks (Coke, Pepsi, and RC Cola), paper towels (Brawny, Bounty, and Scott), cereal (Kellogg’s, General Mills, Post, Kashi, and Jewel), and sports drinks (Gatorade, Powerade, and SoBe). A pretest determined that consumers classify brands in the first four categories (sneakers, knit shirts, jackets, and watches) as primarily symbolic, whereas they classify brands in the remaining four categories (soft drinks, paper towels, cereal, and sport drinks) as primarily functional. Each respondent evaluated brands in four product categories, all either symbolic or functional. The task was identical to the one used in the first experiment and required allocating 100 points across the available brands according to respondents’ perceptions of the brands’ personal relevance. Results We hypothesized that the strength of consumer brand preferences is influenced by the availability of nonbrand means of self-expression and that this effect is more pronounced for symbolic than functional brands. As in the first experiment, we operationalized brand value as (1) the absolute score of the brand perceived to be most personally relevant and (2) the point difference (distance) between the most relevant and second most relevant brand. Each of the 104 respondents evaluated brands in four product categories, which yielded 411 ratings (5 missing data points): 201 in the high self-expression condition and 210 in the low self-expression condition. The data summarized in Table 1 show that respondents who were asked to TABLE 1 The Effect of Self-Expression on Consumer Preferences for Symbolic and Functional Brands (Experiment 2) Self-Expression High Low Brand Type Dependent Variable (Listing Favorites) (Filler Task) Effect Symbolic Personal relevance (absolute) 46.5 63.4 F1, 100 = 22.62 (n = 104, SD = 19.2) (n = 99, SD = 22.7) p < .001 Personal relevance (marginal) 17.6 42.2 F1, 100 = 24.87 (n = 104, SD = 25.6) (n = 99, SD = 33.8) p < .001 Functional Personal relevance (absolute) 51.9 58.9 F1, 100 = 3.90 (n = 97, SD = 18.7) (n = 111, SD = 24.0) p < .05 Personal relevance (marginal) 22.8 33.9 F1, 100 = 5.10 (n = 97, SD = 24.2) (n = 111, SD = 34.5) p < .05 Combined Personal relevance (absolute) 49.1 61.0 F1, 100 = 22.66 (n = 201, SD = 19.1) (n = 210, SD = 23.4) p < .001 Personal relevance (marginal) 20.2 37.8 F1, 100 = 25.97 (n = 201, SD = 25.1) (n = 210, SD = 34.3) p < .001 Notes: Numbers in each cell represent the mean ratings of the brands across all product categories. articulate their preferences before the evaluation task rated the brands as being less relevant both in terms of absolute ratings (M = 49.1 vs. M = 61.0) and marginal ratings (M = 20.2 vs. M = 37.8) than respondents in the control condition. This effect was statistically significant (F1, 100 = 22.66, p < .001, and F1, 100 = 25.97, p < .001, respectively) and consistent across product categories, as revealed by a nonsignificant interaction between the factors of interest and product category (F3, 295 =.56 and F3, 295 =.25, respectively). These findings are consistent with the hypothesis that brand value, as measured by perceived brand relevance, is a function of the availability of nonbrand means of selfexpression. The data further show that the self-expression task had a significant impact on the perceived brand relevance for both symbolic and functional brands. For example, the absolute ratings of brand relevance in the high self-expression and low self-expression conditions were M = 46.5 and M = 63.4 for symbolic brands (F1, 100 = 22.62, p < .001) and M = 51.9 and M = 58.9 for functional brands (F1, 100 = 3.90, p < .05). The marginal ratings of brand relevance revealed the same pattern. More important, the data show that the impact of the self-expression task on perceived brand relevance was a function of brand type. Namely, the self-expression task had a greater impact on brand relevance for symbolic than functional brands, as indicated by both absolute and marginal measures of brand relevance (F1, 100 = 3.96, p < .05, and F1, 100 = 3.83, p < .05). These findings are consistent with the proposition that the identity saturation effects resulting from self-expression are more pronounced for symbolic than functional brands. Discussion The data furnished by Experiment 2 lend further support to the proposition that increasing the prominence of selfexpressive brands that are already a part of a consumer’s identity is likely to weaken his or her future brand preferences. In addition, the data show that this effect can be caused by nonbrand means of self-expression, such as listing favorite books and sports teams, and that it also stems from the self-expressive function of the brands, such that it is more pronounced for symbolic than functional brands. Considered together, the data from the first two experiments offer converging evidence that asking respondents to articulate personally relevant items (both brand and nonbrand) tends to result in greater indifference to brands in subsequent choice tasks. An important issue that merits consideration is that in both experiments we manipulated the level of identity saturation by explicitly prompting respondents to articulate their preferences. This raises the question whether the effects reported in the first two experiments would persist in a scenario in which people achieve identity saturation by simply expressing their preferences. Experiment 3 addresses this question by investigating whether the mere act of evaluating personally relevant brands can weaken preferences for subsequently evaluated brands. This proposition is based on the notion that evaluation itself can serve a self-expressive function (Kim and Drolet 2003; Kim and Sherman 2007) and, thus, that mak72 / Journal of Marketing, May 2011 ing multiple selections in personally relevant categories is likely to lead to identity saturation. The proposition that the mere act of evaluating a brand’s personal relevance can lead to identity saturation is consistent with the data reported in the first experiment. Indeed, the data collected in Experiment 1 show that the brands evaluated later in the set were more likely to be rated lower in terms of personal relevance than brands evaluated earlier in the set—an effect significant both in terms of absolute (F1, 380 = 7.87, p < .01) and relative (F1, 380 = 5.25, p < .05) measures of personal relevance. Although these data are consistent with the notion that the mere act of evaluating personally relevant brands can serve as a means of self-expression, they do not provide conclusive evidence. Because brands were presented to all respondents in the same order, effects attributed to the presentation order were perfectly confounded with brand-specific effects. Furthermore, one could argue that the decrease in brand relevance can be attributed to respondents simply becoming more tired (or ego depleted; see Baumeister, Smart, and Boden 1996). That is, the increase in respondents’ indifference to brands evaluated later in the sequence could have been caused by resource depletion rather than identity saturation. Thus, we predict that a person’s brand preferences are a function of the self-expressive capacity of the previously considered brands, such that merely evaluating a self-expressive brand can lower consumer preferences for subsequently evaluated self-expressive brands. Experiment 3 tests the proposition that the mere act of brand evaluation can serve a self-expressive function, while controlling for brand order and resource depletion effects. Experiment 3 Experiment 3 examines identity saturation effects naturally occurring in the process of product evaluations. In particular, it tests the proposition that the mere act of evaluating a self-expressive brand can weaken consumer preferences for subsequently considered brands. Method In this experiment, we presented respondents with a set of brands in different categories and compared their evaluations of brands considered at both the beginning and the end of the list. To control for potential resource depletion effects, our design employed the following procedure. First, we presented respondents with a list of 15 categories and asked them to identify the five most personally relevant and five least personally relevant ones. Then, respondents were randomly assigned to one of two self-expression conditions. Respondents in the high self-expression condition were given five sets of brands in categories identified by respondents as the most relevant, followed by five sets of brands with average relevance (those not identified either as most or least relevant). In contrast, respondents in the low selfexpression condition were given five sets of brands identified as the least relevant, followed by five sets of brands with average relevance. Because respondents in both conditions evaluated the same number of brands, any differences in their evaluations of the second group of brands (those with average relevance) could be attributed to identity expression effects caused by evaluating the first group of brands. Respondents were 109 students who were randomly assigned to either the high or low self-expression condition. The experiment was conducted online, and the ten decision sets given to each respondent were dynamically generated according to their initial evaluation of the relevance of the available product categories. The evaluation task involved rating the personal relevance of the brands in each product category (100-point allocation task) and measuring the perceived similarity of the available brands (100-point ungraded scale presented to respondents in the form of a slider with the endpoints “very similar” and “very dissimilar”). Results As a manipulation check, we compared the brand relevance and similarity ratings in either the five high-relevance categories in the high self-expression condition or the five lowrelevance categories in the low self-expression condition with the ratings in the five average-relevance categories in both conditions. The data show that the manipulation had the predicted effect in both conditions, such that brands in the average-relevance categories were rated as less relevant than brands in the high-relevance categories and more relevant than brands in the low-relevance categories. In particular, in the high self-expression condition, the brand-relevance ratings in the average-relevance categories were lower than those in the high-relevance categories (absolute ratings: M = 40.1 vs. M = 51.7; F1, 465 = 39.11, p < .001; marginal ratings: M = 16.9 vs. M = 29.6; F1, 465 = 28.73, p < .001). In the low self-expression condition, the brand-relevance ratings in the average-relevance categories were higher than those in the low-relevance categories (absolute ratings: M = 52.5 vs. M = 45.2; F1, 421 = 8.74, p < .005; marginal ratings: M = 30.4 vs. M = 20.3; F1, 418 = 15.21, p < .001). We hypothesized that evaluating brands in personally relevant product categories would serve as a means of selfexpression and, therefore, would weaken preferences for subsequently presented brands. In contrast, we predicted that evaluating brands in categories that were perceived as less relevant would result in a less pronounced self-expression effect and, thus, have little or no impact on subsequent brand preferences. Each of the 109 respondents evaluated brands in ten product categories, which yielded 1054 personal releCompeting for Consumer Identity / 73 vance ratings (552 in the self-expression condition and 502 in the control condition, with 36 missing observations) and 1074 similarity ratings (561 in the self-expression condition and 513 in the control condition). The data summarized in Table 2 show that the ratings of the most personally relevant brand in the second group of brands (the five categories rated as having average personal relevance) were lower for respondents in the high selfexpression condition, who were first asked to evaluate brands in the most relevant product categories, than for respondents in the low self-expression condition, who were first asked to evaluate brands in the least relevant product categories (M = 40.1 vs. M = 52.5). This effect was statistically significant (F1, 106 = 15.22, p < .001). The marginal ratings displayed a similar pattern, indicating that respondents’ preferences were weaker when the evaluation of brands in the average-relevance categories was preceded by evaluation of brands in self-expressive (personally relevant) categories (M = 16.9 vs. M = 30.4; F1, 106 = 11.41, p < .001). The brand similarity ratings further show that respondents in the high self-expression condition perceived the available brands to be more similar to one another than respondents in the low self-expression condition (M = 35.9 vs. M = 45.1; F1, 107 = 7.12, p < .01). All three measures were consistent across all five product categories, as revealed by nonsignificant interactions between the factors of interest and product category (F4, 384 = .80, F4, 381 = 1.00, and F4, 393 = 1.21, respectively). Discussion The data reported in Experiment 3 support the proposition that the mere act of evaluating a brand in a personally relevant product category can serve a self-expressive function, weakening people’s preferences for subsequently evaluated brands. Thus, respondents who were asked to evaluate brands in categories of greater personal relevance perceived the subsequently rated brands to be more similar in their personal relevance. The data from the three experiments reported so far are consistent with the proposition that the self-expressive value of a given brand tends to decrease as the number of alternative brand and nonbrand means of self-expression increases. In particular, we varied the salience of brands that were already a part of consumers’ identity and attributed the resulting changes in brand preferences to the TABLE 2 The Effect of Self-Expressive Product Evaluations on Brand Preferences (Experiment 3) Self-Expression High Low Dependent Variable (High Category Relevance) (Low Category Relevance) Effect Personal relevance (absolute) 40.1 52.5 F1, 106 = 15.22 (n = 272, SD = 22.6) (n = 248, SD = 27.7) p < .001 Personal relevance (marginal) 16.9 30.4 F1, 106 = 11.41 (n = 272, SD = 27.4) (n = 245, SD = 36.6) p < .001 Perceived similarity 35.9 45.1 F1, 107 = 7.12 (n = 276, SD = 24.9) (n = 254, SD = 26.6) p < .01 Notes: Numbers in each cell represent the mean ratings of the five target brands across all product categories. Lower ratings indicate greater perceived similarity of the brands in the category. degree to which consumers had already expressed their identity. An alternative and more direct approach to demonstrating the causal link between the degree of self-expression and the strength of brand preferences is to vary consumers’ need for self-expression. We argue that the decline in the strength of consumer brand preferences stems from a decline in the need for self-expression because of the availability of alternative means of expressing identity. Following this line of logic, we would expect that increasing consumers’ need for self-expression would have the opposite effect of strengthening their brand preferences. One approach to increasing people’s need for selfexpression is to introduce a threat to their self-identity (Berger and Heath 2007; Escalas and Bettman 2005; Wicklund and Gollwitzer 1981). Consistent with this approach, previous research has shown that brands can be valued as a way of making up for deficiencies in a self-concept—a way of symbolically completing the self. For example, MBA students aspiring to be perceived as “successful” business people were more likely to wear business-appropriate attire, including luxury watches and other accessories, than students who had already achieved some objective degree of success (Wicklund and Gollwitzer 1981). Likewise, among a sample of self-identified “committed” tennis players, beginners were much more likely to prefer specific brands of clothing when they played relative to those with several years of experience, who had weaker clothing brand preferences (Braun and Wicklund 1989). In this context, we predict that the presence of a threat to people’s identity will increase their need for self-expression and strengthen their brand preferences, attenuating the identity saturation effect observed in the first three experiments. Therefore, we predict that the impact of the availability of alternative means of self-expression on consumer preferences for self-expressive brands is a function of people’s need to express their identities, such that it will be more pronounced when the need for self-expression is high. We test the validity of this prediction in Experiment 4. Experiment 4 Experiment 4 tests the hypothesis that the impact of alternative means of self-expression on brand preferences is a function of a person’s need for self-expression. In particular, we manipulated people’s need for self-expression by providing feedback that was either consistent or inconsistent with their self-image and examined its impact on subsequent brand preferences. Method Respondents were 122 students recruited to participate in the experiment. They were initially given a set of selfexpressive questions asking them to identify their favorite color, sport teams, music group, hobby, television show, book, subject at school, and movie. Following the initial self-expressive task, respondents were informed that their preferences were either very unique (identity-validation condition) or very similar to those of the other respondents (identity-threat condition). The premise was that respondents 74 / Journal of Marketing, May 2011 in the identity-threat condition would have a greater need to express their identity than those in the identity-validation condition (Berger and Heath 2007; Brewer 1991; Escalas and Bettman 2005; Rucker and Galinsky 2008). Next, all respondents were given the same set of brandchoice questions in four categories: sunglasses (Oakley, Foster Grant, Revo, Fossil, and Ray-Ban), knit shirts (Polo, Nautica, Abercrombie & Fitch, Lacoste, and Izod), pens (Montblanc, Parker, uni-ball, Pilot, and Office Depot), and jackets (The North Face, Timberland, Patagonia, Columbia, and Mountain Hardwear). In each category, respondents were asked to indicate (1) the personal relevance of each brand by allocating 100 points across all brands, (2) the degree to which the available brands were similar to one another (using a 100-point nongraded slider with the endpoints “very similar” and “very dissimilar”), and (3) the amount of money they were willing to pay for each of the available brands. These three measures served as the operationalization of the strength of brand preferences. The first two measures were similar to those used in the first three experiments. We measured the impact of the need for self-expression on respondents’ willingness to pay by comparing the amount of money respondents were willing to pay for their most preferred brand and the premium they were willing to pay for their most preferred brand over their second most preferred brand. Results We argue that the strength of people’s brand preferences is a function of the need for self-expression. In particular, we hypothesize that respondents in the identity-threat condition were likely to display stronger preferences when evaluating the individual brands than those in the identity-validation condition. The data summarized in Table 3 confirm our hypotheses, showing that respondents in the identity-threat condition displayed stronger brand preferences, as indicated by the dispersion of their personal relevance brand rankings, brand similarity ratings, and willingness to pay for their most preferred brand. To illustrate, respondents who were told that their preferences were similar to those of others (identity-threat condition) rated their most relevant brand more highly than those who were told that their preferences were unique (identity-validation condition): (M = 54.1 vs. M = 42.1; F1, 120 = 21.70, p < .001). The effect was consistent across categories, although it seemed to be stronger in some categories, as indicated by a significant interaction between factors of interest and product category (F3, 348 = 4.20, p < .01). The willingness-to-pay data summarized in Table 4 show that varying the need for self-expression had a significant impact on respondents’ willingness to pay for the available brands. For example, respondents in the identity-threat condition were willing to pay on average $133.25 for a sport jacket, whereas those in the identity-validation condition were willing to pay only $99.20. Furthermore, respondents in the identity-threat condition were willing to pay an average premium of $43.33 for their most preferred brand (over what they would pay for their second most preferred brand), whereas those in the identity-validation condition were only willing to pay an average premium of $18.48. Discussion The data reported in this experiment lend further support to the notion that the strength of people’s brand preferences is a function of their need for self-expression. In particular, we show that decreasing the need for self-expression (e.g., by validating people’s unique identity) tends to weaken their brand preferences, whereas increasing their need for selfexpression (e.g., by threatening their identity) has the opposite effect of strengthening their brand preferences. The data show convergence across all three of the experiment’s measures—personal brand relevance, perceived brand similarity, and willingness to pay—which enhances the validity of the observed effects. Because personal relevance, perceived similarity, and willingness to pay represent different aspects of consumers’ brand preferences, the findings suggest that the observed impact of identity saturation on brand value is not contingent on the instrument used but rather reflects a more fundamental pattern of consumer information processing. The data from this experiment lend further support to the notion that a person’s need for self-expression can be satiated through various means. In particular, it could be argued that the identity saturation manipulation used in some of the previous experiments might have inadvertently changed the way people evaluated different brands. Thus, listing items that are most identity relevant might have made subsequently listed brands less identity relevant by Competing for Consumer Identity / 75 comparison. This argument might apply to the first two experiments, which ask respondents to generate identityrelevant lists of items, and to a lesser degree to the third experiment, in which identity saturation follows from respondents’ evaluation of self-relevant brands. Experiment 4 controls more directly for the possibility that people differentially weight subsequently evaluated brands. Indeed, in this experiment, all respondents were asked to articulate their preferences, so any change in the importance of the sequentially evaluated brands would have been consistent across conditions. In general, we argue that because people’s need for selfexpression is finite, the value they derive from brands decreases as their need for self-expression is satiated. The first four experiments tested this proposition by increasing the salience of self-expressive aspects of respondents’ identity, such as their favorite brands, music, sports teams, and hobbies. Rather than increasing the prominence of an already established aspect of a person’s identity, an alternative approach to examine the boundaries of self-expression is to actively engage people in a self-expressive activity. Indeed, our theorizing predicts that engaging in self-expressive acts—including decorating a house, creating a movie for YouTube, and customizing a product—would temporarily satiate people’s need for self-expression, which in turn is likely to weaken their preferences for subsequently evaluated brands. In this context, we argue that people’s brand TABLE 3 The Effect of Self-Expression and Identity Threat on Brand Preferences (Experiment 4) Need for Self-Expression High Low Dependent Variable (Identity Threat) (Identity Validation) Effect Personal relevance (absolute) 54.1 42.1 F1, 120 = 21.70 (n = 231, SD = 23.2) (n = 245, SD = 22.7) p < .001 Personal relevance (marginal) 28.3 17.0 F1, 120 = 17.21 (n = 231, SD = 32.3) (n = 245, SD = 25.8) p < .001 Perceived similarity 45.5 38.1 F1, 120 = 7.84 (n = 231, SD = 24.4) (n = 245, SD = 22.8) p < .01 Willingness to pay for $82.99 $60.25 F1, 119 = 6.86 the most preferred brand (n = 229, SD = 79.5) (n = 238, SD = 66.9) p < .01 Premium willing to pay for $25.83 $16.75 F1, 119 = 5.42 the most preferred brand (n = 229, SD = 41.2) (n = 238, SD = 35.1) p < .05 Notes: Numbers in each cell represent the mean brand relevance, similarity, and willingness to pay across all product categories. Lower ratings indicate greater perceived similarity of the brands in the category. TABLE 4 The Effect of Self-Expression and Identity Threat on Consumer Willingness to Pay (Experiment 4) Need for Self-Expression High Low Product Category (Identity Threat) $ (Identity Validation) $ Effect Jackets 133.26 (43.33) 99.20 (18.48) 25.6 (57.4) Pens 45.31 (20.46) 26.79 (16.82) 40.9 (17.8) Knit shirts 66.57 (17.59) 48.64 (9.39) 26.9 (46.6) Sunglasses 87.12 (22.10) 67.32 (22.38) 22.7 (–1.3) Overall 82.99 (25.83) 60.25 (16.75) 27.4 (35.1) Notes: Numbers in each cell indicate respondents’ willingness to pay for the most preferred item. The price premium over the next most preferred brand is in parentheses. preferences are a function of the extent to which they have engaged in self-expressive behavioral acts, such that selfexpressive actions tend to decrease consumer preferences for subsequently evaluated brands. We test this prediction in the following experiment. Experiment 5 Experiment 5 aims to provide further evidence for the proposition that people’s need for self-expression through brands is finite. We do so by engaging respondents in selfexpressive behavior and examining its impact on their subsequent reliance on brands to express their identity. Method Respondents, 87 adult consumers participating in an online research panel, completed two ostensibly unrelated tasks. The first task involved designing a T-shirt using Blue Cotton’s online T-shirt design studio (BlueCotton.com). After selecting a T-shirt color, participants were able to customize their T-shirts by adding text, numbers, shapes, and pictures from an extensive selection of clip art organized by category (e.g., occupations, music, animals). Design elements could be added to both the front and back of the shirt. Participants could customize the color, size, location, and orientation of each element and were even able to add visual effects, such as distressing. To manipulate people’s need for self-expression, some of the participants were asked to design a shirt for themselves (high self-expression condition), whereas others were asked to design a T-shirt for a parent (low self-expression condition). Following the product customization task, all respondents were given a set of brand evaluation questions in six categories: sunglasses (Oakley, Foster Grant, Revo, Fossil, and Ray-Ban), knit shirts (Polo, Nautica, Abercrombie & Fitch, Lacoste, and Izod), backpacks (The North Face, JanSport, Columbia, Mountain Hardware, and Timbuk2), pens (Montblanc, Parker, uni-ball, Pilot, and Office Depot), jackets (The North Face, Timberland, Patagonia, Columbia, and Mountain Hardwear), and watches (Movado, Swatch, Rolex, Seiko, and Omega). In each category, respondents indicated (1) the personal relevance of each brand by allocating 100 points across all brands, (2) the degree to which the available brands were similar to one another, and (3) the amount of money they were willing to pay for each of the available brands. These three questions were identical to the ones used in Experiment 4. As an additional measure of brand relevance, we examined the closeness of the self-reported relationship people had to their most preferred brand from the available set. To that end, we developed a specialized scale, which we refer to as the Brand Closeness Scale. Derived from the Inclusion of Other in the Self scale (Aron, Aron, and Smollan 1992), this scale provides a more direct measure of the impact of identity saturation on people’s relationship with the subsequently presented brands. The use of this scale to measure a brand’s personal relevance is further supported by the growing body of research suggesting that brand relationships bear striking similarities to interpersonal relationships (Aaker, Fournier, and Brasel 2004; Fournier 1998). The 76 / Journal of Marketing, May 2011 Brand Closeness Scale consists of seven pairs of partially overlapping circles. Respondents were asked to select the pair that best represented their relationship with their most preferred brand in each category. After completing the brand evaluations, participants were asked a set of three questions designed to measure their need for self-expression. Respondents used a fivepoint scale (1 = “strongly disagree” to 5 = “strongly agree”) to assess the following items: “I would like to be perceived as different from the general population”; “I often purchase products that let me express my uniqueness”; and “It is important for me to be able to express my identity.” We used participants’ responses on these scales to check the effectiveness of the self-expression manipulation. Thus, we expected respondents in the high self-expression condition (designing a T-shirt for self) to display lower subsequent need for self-expression compared with respondents in the low self-expression condition (designing a T-shirt for a parent). The manipulation check was conducted at the end of the experiment to ensure that it did not influence people’s responses to the key dependent measures. Results To test the effectiveness of the manipulation procedure, we examined participants’ responses on the three-item scale measuring the need for self-expression. The responses were highly correlated (a = .76), so we averaged the three responses to form a single indicator of the need for selfexpression. The manipulation check data revealed that participants in the high self-expression condition, who designed a T-shirt for themselves, displayed lower subsequent need for self-expression (M = 3.24) than those in the low selfexpression condition, who designed a T-shirt for a parent (M = 3.73; F1, 80 = 6.57, p < .05). These findings indicate that the manipulation procedure (customizing a T-shirt for self vs. a parent) was successful in creating different levels of identity saturation and thus different levels of need for self-expression among respondents. We theorized that the strength of people’s brand preferences is a function of their need for self-expression. In particular, we expected that respondents in the high selfexpression condition, who designed a T-shirt for themselves, were likely to display weaker subsequent brand preferences than those in the low self-expression condition, who designed a T-shirt for a parent. The data summarized in Table 5 are consistent with this prediction. In particular, the data show that (compared with those in the low self-expression condition) respondents in the high self-expression condition (1) regarded their most preferred brand as having lower personal relevance, in terms of both absolute and marginal (the difference between the most preferred and the next most preferred brand) ratings, (2) perceived their relationship with the preferred brand to be more distant, (3) perceived the brands in the choice set to be more similar to one another, and (4) were willing to pay less for their preferred brand in terms of both the absolute dollar amount and the premium over the next most preferred brand. To illustrate, respondents in the high self-expression condition rated their most relevant brand lower than those in the low self-expression condition in terms of both their absolute (M = 38.5 vs. M = 55.6; F1, 412 = 23.50, p < .001) and marginal (M = 15.0 vs. M = 31.2; F1, 85 = 17.00, p < .001) ratings. Furthermore, participants in the high selfexpression condition rated their relationship with the most preferred brand to be more distant than those in the low self-expression condition (M = 2.7 vs. M = 3.2; F1, 85 = 4.98, p < .05). In the same vein, respondents in the high self-expression condition perceived the available brands to be more similar than those in the low self-expression condition (M = 33.7 vs. M = 46.6; F1, 85 = 20.76, p < .001). The willingness-to-pay data further show that engaging in a self-expressive act has a significant impact on respondents’ willingness to pay for their most preferred brand: Respondents in the high self-expression condition were willing to pay on average $73.41 for their most preferred brands, whereas those in the low self-expression condition were willing to pay on average $111.26 (F1, 85 = 5.89, p < .05). Furthermore, respondents in the high self-expression condition were willing to pay only a $27.08 premium for their most preferred brand, significantly less than those in the low self-expression condition, who were willing to pay a premium of $57.33 (F1, 85 = 6.95, p < .05). These findings are consistent with the experimental predictions. Competing for Consumer Identity / 77 Because the willingness-to-pay-data were category specific (e.g., across conditions, respondents were willing to pay more for a jacket than a pen), we report the per-category effects in Table 6 by category. To illustrate, respondents in the high self-expression condition were willing to pay on average $58.00 for a pair of sunglasses, whereas those in the low self-expression condition were willing to pay $72.57. Furthermore, respondents in the high self-expression condition were willing to pay an average premium of $17.98 for their most preferred brand over their second most preferred brand, whereas those in the low self-expression condition were willing to pay an average premium of $29.80. There was a significant main effect of product category on the willingness-to-pay measures (absolute willingness to pay: F5, 508 = 23.94, p < .001; marginal willingness to pay: F5, 508 = 11.08, p < .001). The effects were consistent across product categories, as indicated by the nonsignificant interactions between product category and the experimental manipulation (absolute willingness to pay: F5, 508 = 2.11; marginal willingness to pay: F5, 508 = 1.20). Discussion The data reported in this experiment provide further support for the notion that the strength of people’s brand preferTABLE 5 The Effect of Self-Expression Product Customization on Brand Preferences (Experiment 5) Need for Self-Expression High Low Dependent Variable (Design T-Shirt for Self) (Design T-Shirt for Other) Effect Personal relevance (absolute) 38.5 55.6 F1, 85 = 23.50 (n = 250, SD = 20.7) (n = 261, SD = 25.4) p < .001 Personal relevance (marginal) 15.0 31.2 F1, 85 = 17.00 (n = 248, SD = 22.6) (n = 260, SD = 33.5) p < .001 Closeness of relationship 2.7 3.2 F1, 85 = 4.98 with most preferred brand (n = 248, SD = 1.5) (n = 261, SD = 1.6) p < .05 Perceived similarity 33.7 46.6 F1, 85 = 20.76 (n = 248, SD = 20.7) (n = 261, SD = 24.4) p < .001 Willingness to pay for $73.41 $111.26 F1, 85 = 5.89 the most preferred brand (n = 248, SD = 125.0) (n = 261, SD = 154.7) p < .05 Premium willing to pay for $27.08 $57.33 F1, 85 = 6.95 the most preferred brand (n = 248, SD = 108.8) (n = 261, SD = 122.6) p < .01 Notes: Numbers in each cell represent the mean brand relevance, relationship closeness, similarity, and willingness to pay across all product categories. Higher relationship closeness ratings indicate greater perceived closeness between the self and the most preferred brand. Lower similarity ratings indicate greater perceived similarity of the brands in the category. TABLE 6 The Effect of Self-Expression Product Customization on Consumer Willingness to Pay (Experiment 5) Self-Expression High Low Product Category (Design T-Shirt for Self) $ (Design T-Shirt for Other) $ Effect Sunglasses 58.00 (17.98) 72.57 (29.80) 20.1 (39.7) Knit shirts 40.19 (8.93) 57.61 (13.80) 30.1 (35.3) Backpacks 79.52 (24.17) 82.07 (31.16) 3.1 (22.4) Pens 20.12 (11.81) 74.67 (60.21) 73.1 (80.4) Jackets 87.65 (16.63) 117.33 (56.91) 25.3 (70.8) Watches 159.75 (85.25) 266.12(154.37) 40.0 (44.8) Overall 73.41 (27.08) 111.26 (57.33) 34.0 (52.8) Notes: Numbers in each cell indicate respondents’ willingness to pay for the most preferred item. The price premium over the next most preferred brand is in parentheses. ences is a function of their identity saturation. In particular, we show that self-expressive behaviors tend to weaken subsequent brand preferences. This finding extends the results reported in Experiments 1–4 by demonstrating the impact of identity saturation on brand preferences as a direct result of self-expressive acts. We document these effects across all three key dependent variables used in the first four experiments: personal brand relevance, perceived brand similarity, and willingness to pay. This experiment provides further support to our theory by using a measure of a brand’s proximity to the self as an additional, and perhaps more direct, measure of a brand’s personal relevance. In this context, we show that identity saturation is likely to lead to a greater perceived distance between a person’s self and his or her preferred brands. Consistent with the proposition that the need to express the self is finite, we show that participants engaging in highly self-expressive acts were less likely to incorporate preferred brands into their self-concept. The data reported in this experiment lend further support to the notion that self-expression can be satiated through any number of means. The previous studies reveal that the need for self-expression can be saturated by generating self-relevant information from memory (Experiments 1, 2, and 4) and even by simply evaluating a list of brands (Experiment 3). Building on these findings, Experiment 5 shows that self-expressive behaviors, such as product customization, can satiate a person’s need for self-expression, weakening brand preferences. The use of an actual productcustomization task provides further evidence for the external validity of the observed identity saturation effect. 78 / Journal of Marketing, May 2011 General Discussion Theoretical Contribution This research offers empirical evidence that a person’s need for self-expression is finite and the preference for selfexpressive brands is contingent on the availability of alternative means of self-expression. In particular, we show that increasing the prominence of self-expressive brands that are already a part of a consumer’s identity is likely to weaken their brand preferences in the immediate future. We further show that this effect is not limited to brands in the same product category but can be activated by identifying selfexpressive brands in unrelated product categories (Experiment 1). We show that nonbrand means of self-expression, such as listing favorite books or television shows, can also cause identity saturation and are likely to have a greater impact on symbolic than functional brands (Experiment 2). In addition, our findings reveal that the mere act of evaluating personally relevant brands can serve a self-expressive function, weakening consumers’ preferences for subsequently evaluated brands (Experiment 3). We further show that decreasing people’s need for self-expression (e.g., by validating their unique identity) tends to weaken their brand preferences, whereas increasing their need for self-expression (e.g., by threatening their identity) has the opposite effect of strengthening their brand preferences (Experiment 4). Finally, we show that self-expressive behavioral acts, such as product customization, can lead to identity saturation, weakening consumers’ brand preferences (Experiment 5). Figure 1 illustrates the overall experimental paradigm and the contribution of individual experiments to testing our FIGURE 1 Overview of the Experiments Experiments 1, 3 Experiments 2, 4 Experiment 5 Experiments 1–5 Experiments 3–5 Experiments 4, 5 Experiment 1 Experiment 4 Initial Need for Self-Expression Act of SelfExpression Brand Preferences SelfRepresentation Identity Threat Subsequent Need for SelfExpression Self-Expression Through Unrelated Brands Self-Expression Through Nonbrand Items Self-Expression Through Behavioral Acts A Brand’s Personal Relevance A Brand’s Similarity to Other Brands Willingness to Pay for Preferred Brand Notes: Boxes in bold represent the hypothesized process. The nonbold boxes represent empirical operationalizations of the independent and dependent variables in each experiment. Identity saturation occurs when a self-expressive act results in a lower subsequent need for self-expression. ˛ Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô ˝ Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô Ô ¸ ¸ Ô Ô Ô Ô Ô Ô Ô Ô ˝ Ô Ô Ô Ô Ô Ô Ô Ô ˛ ¸ Ô Ô Ô Ô Ô Ô Ô Ô ˝ Ô Ô Ô Ô Ô Ô Ô Ô ˛ theoretical predictions. We measure respondents’ need for self-expression by introducing a self-expression manipulation using unrelated self-expressive brands (Experiments 1 and 3), nonbrand self-expressive items (Experiments 2 and 4), and self-expressive behavioral acts (Experiment 5). We also measure (Experiment 1) and manipulate (Experiment 4) the need for self-expression to examine the impact of identity saturation on brand preferences. We document the impact of self-expression on the strength of brand preferences using three complementary measures: personal brand relevance (Experiments 1–5), brand similarity (Experiments 3–5), and willingness to pay (Experiments 4 and 5). The convergence of the data across different manipulations of the need for self-expression and different measures of brand preferences indicates that the observed effects are not contingent on the particulars of the experimental manipulations but rather reflect a more fundamental pattern of consumer information processing. From a theoretical standpoint, our research contributes to the self-identity literature by advancing the notion that people’s need for self-expression is finite and that increasing the variety of means of self-expression can lower the self-expressive capacity of each individual item. We further show that the need for self-expression can be satiated by means of different product categories and across consumption occasions. Our research further adds to the literature on satiation by documenting that satiation can occur across domains. Thus, prior research has argued that past experience influences current evaluations when both are in the same category (Brown 1953; Epstein et al. 1993; Kahneman and Miller 1986) and, as a result, satiation is domain specific. For example, previous research has shown that in the food domain, satiation occurs within similar food categories rather than across all foods (Rolls, Rowe, and Rolls 1982). In contrast, we show that needs that are more fundamental in nature, such as the need for self-expression, can be satiated by means that extend beyond a specific category. We also show that satiation is not limited to the sensory aspect of the product experience (Inman 2001; Johnson and Vickers 1993) but extends to nonsensory features like brand meaning. Sensory-specific satiety has been attributed to habituation, whereby frequent exposure to a stimulus decreases the magnitude of a person’s response to it (Berlyne 1971; McSweeney and Swindell 1999; Raynor and Epstein 2001). Our research adds to this literature by documenting that satiation can occur on nonsensory dimensions, such as a person’s need for self-expression. Managerial Implications We argue that managers could benefit from expanding their understanding of the scope of brand competition to include the idea that brands can potentially compete across categories to become a part of a person’s identity. Conventional wisdom suggests that the strength of consumer brand preferences is a function of the relative strength of other brands in the same category. In contrast, we propose that in the case of brands fulfilling a self-expressive need, consumer preferences are a function of the strength of the brands not only in the same category but also across categories and can Competing for Consumer Identity / 79 include even nonbrand means of self-expression. Thus, in the domain of self-expression, a cereal brand competes not only with other cereal brands but also with all brands with a self-expressive component, including apparel (e.g., Ralph Lauren), electronics (e.g., Apple), and sporting goods (e.g., Nike). Our theorizing offers important insights into the viability of lifestyle positioning as a brand differentiation strategy. Lifestyle positioning has become an increasingly common approach among managers, especially in commodity categories in which functional differences are difficult to maintain. Thus, in addition to brands positioned as lifestyle, such as Ralph Lauren, Abercrombie & Fitch, and Martha Stewart, several well-established brands, such as Gillette, Dove, Montblanc, Oakley, and Quiksilver, have transitioned from being performance focused to lifestyle oriented. To many managers, lifestyle brands seem to offer a way of breaking free of the cutthroat competition within a category by connecting with consumers on a more personal level. However, we argue that the open vistas of lifestyle branding may be an illusion: Managers may be trading fierce within-category functional competition for fierce across-category symbolic competition whereby all self-expressive brands could end up competing with one another. Moreover, these brands might be competing with an increasing number of nonbrand self-expressive items, in addition to self-expressive behaviors that include social media websites (e.g., Facebook), product customization websites (e.g., Nikeid.com), and even self-expressive self-customization websites (e.g., SecondLife. com). Thus, by switching from functional branding to lifestyle branding, managers might be setting themselves up for even stronger competition for a share of consumers’ identity. In general, identity saturation can influence brand preferences in two ways. First, exposure to a self-expressive brand can lead to a short-term saturation of the need to selfexpress, thus decreasing the strength of a consumer’s preference for additional self-expressive brands. For example, a consumer who has purchased a self-expressive brand is less likely to purchase or less willing to pay a premium for another self-expressive brand. In addition, cumulative exposure to self-expressive brands over time could decrease consumers’ willingness to relate to additional self-expressive brands. For example, the adoption of a self-expressive brand (e.g., Apple, Giorgio Armani, Porsche) might satiate a consumer’s further need for self-expression, thus decreasing preferences for additional self-expressive brands. Although both the short- and long-term impact of identity saturation on brand preferences is important to companies, investigating how the change in short-term preferences translates into longer-term shifts in consumer preferences merits special attention. Indeed, while the immediate impact of identity saturation on brand preferences is likely to influence consumer choices in the short term, the longterm impact of identity saturation is likely to influence the degree to which a brand is internalized by consumers, becoming an integral part of their identities. The latter is also important because it is typically viewed as the key source of brand power and brand equity. Our research empirically examines the short-term effect of identity saturation on brand preferences. In this context, it offers guidance to marketing managers in many of their decisions regarding the immediate settings in which customers interact with brands. For example, this research offers an additional point of consideration when deciding where to locate a retail store. Should the store be located near other stores that could remind consumers of other preferred brands? Although there are benefits to locating in a busy shopping mall (e.g., customer traffic), our research suggests that surrounding consumers with salient reminders of other brands they value may decrease their preference and willingness to pay for any particular brand. Thus, although a store located next to an Apple store may benefit from increased traffic, it may also suffer from consumers whose need for self-expression has been satiated by whatever self-expressive gadget Apple has on display. Our research also has implications for managing consumers’ point-of-purchase behavior. Our data suggest that marketers should exercise caution when engaging customers in selfexpressive activities near the point of purchase, because such activities can lead to counterproductive outcomes, weakening consumers’ brand preferences. The finding that brand preferences are a function of the strength of brands in unrelated product categories and nonbrand means of self-expression has important implications for understanding and measuring the monetary value of the brand to a company. Most of the existing brand valuation methods incorporate the impact of the competition on the strength of each individual brand. However, when estimating the impact of the competition on the equity of a given brand, existing brand valuation methods typically define competition as a domain-specific phenomenon in which brands are primarily competing with other brands within the same product category. None of the existing methods for estimating brand equity take into account the effect that the self-expressive nature of brands in unrelated product categories can have, let alone the effect that the availability of nonbrand means of self-expression may have. This narrow definition of the competitive landscape is likely to lead to overestimation of brand equity. Incorporating the impact of a broader set of self-expressive means, which can have both short- and long-term effects, can contribute to more precise brand equity estimation and enable companies to build stronger brands by more effectively managing their self-expressive function. Thus, although this research empirically examines only the short-term effects of identity saturation, this effect might also have some long-term implications for managing brand equity that could be explored by further research. Further Research Our findings suggest multiple venues for further research. For example, researchers could investigate the duration of the identity saturation effect reported in this research. We examined the impact of identity saturation on brand preferences in the context of a single consumption episode without explicitly investigating the long-term dynamics of this effect. Thus, it is possible that the identity saturation observed in our experiments is transient in nature and will 80 / Journal of Marketing, May 2011 fade away over time. For example, prior research has shown that in sensory-specific domains such as food consumption, satiation effects fade away within a little more than an hour (Hetherington, Rolls, and Burley 1989). People’s ability to “recover” from satiation with the passage of time has been documented in other domains as well (McSweeney and Swindell 1999; Thompson and Spencer 1966). This timebound view of satiation is consistent with a model in which satiation is tied to depletion, whereby the satiated need is replenished over time (McAlister 1982; Vohs et al. 2008). Another promising area for further research involves investigating the individual-specific factors that may moderate identity saturation. Factors that are likely to influence consumers’ need for self-expression and the extent to which identity saturation will influence their brand preferences include the degree to which people strive to differentiate themselves from others and the degree to which they tend to rely on brands to self-express. In addition to investigating individual differences, further research might examine the dynamics of the need for self-expression throughout a person’s lifetime. In this context, it is possible that the need for self-expression through brands is correlated with the formation of a self-concept, such that it is more pronounced during earlier stages of a person’s life cycle and is relatively less prominent in the later stages. Moreover, it could be argued that brands adopted during the formation of a person’s self-concept are more likely to have a greater selfexpressive role and be more difficult to replace compared with brands adopted during later stages of the life cycle. From a longitudinal perspective, we could further argue that the need for self-expression, though finite, remains more or less unchanged over time. This argument implies that although the number of self-relevant brands is likely to decrease over time, brand preferences for the most relevant brands are likely to strengthen over time. Thus, the need for self-expression might remain constant (and finite) over time but be fulfilled by fewer brands during the more advanced stages of a person’s life cycle. In this context, investigating the long-term changes in consumers’ need for self-expression over time and the boundaries to self-expression imposed by its finite nature can shed light on the interplay between short- and long-term effects of identity saturation on brand preferences. Another venue for further research involves investigating the nature of the underlying need for self-expression through brands. In particular, brands can serve at least three self-expressive goals: to identify, to differentiate, and to assimilate. To illustrate, one consumer might adopt the Harley-Davidson brand (e.g., by purchasing a HarleyDavidson or wearing Harley-Davidson apparel) to identify himself with the free spirit that Harley-Davidson stands for, whereas another consumer might adopt the brand to differentiate herself from others by signaling nonconformity, and yet another consumer might adopt this brand to signal his belonging to the Harley-Davidson community. These three self-expressive motivations might also differ in the degree to which the self-expression is public or private. Indeed, when the goal is to identify, self-expression does not necessarily have to be public; it can be self-serving. In contrast, when the underlying motivation is to differentiate or assim- ilate, self-expression is more likely to be public and focused on a person’s social surroundings. Investigating the impact of self-expression on brand preferences as a function of the underlying goals (self-identification, differentiation, or assimilation) and the public nature of self-expression is a promising area for further research. REFERENCES Aaker, Jennifer L. (1997), “Dimensions of Brand Personality,” Journal of Marketing Research, 34 (August), 347–56. ———, Susan Fournier, and S. Adam Brasel (2004), “When Good Brands Do Bad,” Journal of Consumer Research, 31 (1), 1–16. Aggarwal, Pankaj (2004), “The Effects of Brand Relationship Norms on Consumer Attitudes and Behavior,” Journal of Consumer Research, 31 (March), 87–101. Akerlof, George A. and Rachel E. Kranton (2000), “Economics and Identity,” Quarterly Journal of Economics, 115 (3), 715–53. Amaldoss, Wilfred and Sanjay Jain (2005), “Conspicuous Consumption and Sophisticated Thinking,” Management Science, 51 (10), 1449–66. Aron, Arthur, Elaine N. Aron, and D. Smollan (1992), “Inclusion of Other in the Self Scale and the Structure of Interpersonal Closeness,” Journal of Personality & Social Psychology, 63 (4), 596–612. Baumeister, Roy F., Laura Smart, and Joseph M. Boden (1996), “Relation of Threatened Egotism to Violence and Aggression: The Dark Side of High Self-Esteem,” Psychological Review, 103 (January), 5–33. Belk, Russell W. (1988), “Possessions and the Extended Self,” Journal of Consumer Research, 15 (2), 139–68. Berger, Jonah and Chip Heath (2007), “Where Consumers Diverge from Others: Identity Signaling and Product Domains,” Journal of Consumer Research, 34 (June), 121–34. ——— and Morgan Ward (2010), “Subtle Signals and Inconspicuous Consumption,” Journal of Consumer Research, 37 (December), 555–69. Berlyne, Daniel E. (1971), Aesthetics and Psychobiology. New York: Appleton-Century-Crofts. Bodner, Ronit and Drazen Prelec (2003), “Self-Signaling and Diagnostic Utility in Everyday Decision Making,” in The Psychology of Economic Decisions, Isabelle Brocas and Juan D. Carrillo, eds. Oxford: Oxford University Press, 89–104. Braun, Ottmar L. and Robert A. Wicklund (1989), “Psychological Antecedents of Conspicuous Consumption,” Journal of Economic Psychology, 10 (2), 161–87. Brendl, C. Miguel, Arthur B. Markman, and Claude Messner (2003), “The Devaluation Effect: Activating a Need Devalues Unrelated Objects,” Journal of Consumer Research, 29 (March), 463–73. Brewer, Marilynn B. (1991), “The Social Self: On Being the Same and Different at the Same Time,” Personality and Social Psychology Bulletin, 17 (October), 475–82. Broniarczyk, Susan M. and Joseph W. Alba (1994), “The Importance of the Brand in Brand Extension,” Journal of Marketing Research 31 (May), 214–28. Brown, Donald R. (1953), “Stimulus-Similarity and the Anchoring of Subjective Scales,” American Journal of Psychology, 66 (2), 199–214. Bruner, Jerome S. and Cecile C. Goodman (1947), “Value and Need as Organizing Factors in Perception,” Journal of Abnormal and Social Psychology, 42 (1), 33–44. Dolich, Ira J. (1969), “Congruence Relationships Between Self Images and Product Brands,” Journal of Marketing Research, 6 (February), 80–84. Competing for Consumer Identity / 81 Dunning, David (2005), Self-Insight: Roadblocks and Detours on the Path to Knowing Thyself. New York: Psychology Press. Epstein, Leonard H., Anthony R. Caggiula, Joshua S. Rodefer, Lucene Wisniewski, and Shari L. Mitchell (1993), “The Effects of Calories and Taste on Habituation of the Human Salivary Response,” Addictive Behaviors, 18 (2), 179–85. Escalas, Jennifer Edson and James R. Bettman (2005), “SelfConstrual, Reference Groups, and Brand Meaning,” Journal of Consumer Research, 32 (3), 378–89. Fournier, Susan (1998), “Consumers and Their Brands: Developing Relationship Theory in Consumer Research,” Journal of Consumer Research, 24 (March), 343–73. Gardner, Burleigh B. and Sidney J. Levy (1955), “The Product and the Brand,” Harvard Business Review, 33 (2), 33–39. Hetherington, Marion, Barbara J. Rolls, and Victoria J. Burley (1989), “The Time Course of Sensory-Specific Satiety,” Appetite, 12 (February), 57–68. Inman, J. Jeffrey (2001), “The Role of Sensory-Specific Satiety in Attribute-Level Variety Seeking,” Journal of Consumer Research, 28 (June), 105–120. Johnson, Jill and Zata Vickers (1993), “Effects of Flavor and Macronutrient Composition of Food Servings on Liking, Hunger and Subsequent Intake,” Appetite, 21 (August), 25–39. Kahn, Barbara E. and Brian Wansink (2004), “The Influence of Assortment Structure on Perceived Variety and Consumption Quantities,” Journal of Consumer Research, 30 (March), 519–33. Kahneman, Daniel and Dale T. Miller (1986), “Norm Theory: Comparing Reality to Its Alternatives,” Psychological Review, 93 (April), 136–53. Katz, Daniel (1960), “The Functional Approach to the Study of Attitudes,” Public Opinion Quarterly, 24 (2), 163–204. Keller, Kevin Lane (2007), Strategic Brand Management: Building, Measuring and Managing Brand Equity, 3d ed. Upper Saddle River, NJ: Prentice Hall. Kim, Heejung S. and Aimee Drolet (2003), “Choice and SelfExpression: A Cultural Analysis of Variety-Seeking,” Journal of Personality and Social Psychology, 85 (August), 373–82. ——— and David K. Sherman (2007), “Express Yourself: Culture and the Effect of Self-Expression on Choice,” Journal of Personality and Social Psychology, 92 (January), 1–11. Kleine, Robert E., Susan Schultz Kleine, and Jerome B. Kernan (1993), “Mundane Consumption and the Self,” Journal of Consumer Psychology, 2 (3), 209–235. Lattin, James M. and Leigh McAlister (1985), “Using a VarietySeeking Model to Identify Substitute and Complementary Relationships Among Competing Products,” Journal of Marketing Research, 22 (August), 330–39. Levy, Sidney J. (1959), “Symbols for Sale,” Harvard Business Review, 37 (4), 117–24. Lewin, Kurt (1935), A Dynamic Theory of Personality. New York: McGraw-Hill. Loewenstein, George (1999), “Because It Is There: The Challenge of Mountaineering ... For Utility Theory,” Kyklos, 52 (3), 315–43. Markman, Arthur B. and C. Miguel Brendl (2000), “The Influence of Goals on Value and Choice,” The Psychology of Learning and Motivation, 39, 97–128. McAlister, Leigh (1982), “A Dynamic Attribute Satiation Model of Variety-Seeking Behavior,” Journal of Consumer Research, 9 (September), 141–50. McSweeney, Frances K. and Samantha Swindell (1999), “GeneralProcess Theories of Motivation Revisited: The Role of Habituation,” Psychological Bulletin, 125 (July), 437–57. Nelson, Lief D. and Tom Meyvis (2008), “Interrupted Consumption: Disrupting Adaptation to Hedonic Experiences,” Journal of Marketing Research, 45 (December), 654–64. ———, ———, and Jeff Galak (2009), “Enhancing the TelevisionViewing Experience Through Commercial Interruptions,” Journal of Consumer Research, 36 (August), 160–72. Ratner, Rebecca K., Barbara E. Kahn, and Daniel Kahneman (1999), “Choosing Less-Preferred Experiences for the Sake of Variety,” Journal of Consumer Research, 26 (June), 1–15. Raynor, Hollie A. and Leonard H. Epstein (2001), “Dietary Variety, Energy Regulation, and Obesity,” Psychological Bulletin, 127 (May), 325–41. Redden, Joseph P. (2008), “Reducing Satiation: The Role of Categorization Level,” Journal of Consumer Research, 34 (February), 624–34. Richins, Marsha L. (1994), “Valuing Things: The Public and Private Meanings of Possessions,” Journal of Consumer Research, 21 (December), 504–521. Rogers, Carl R. (1947), “Some Observations on the Organization of Personality,” American Psychologist, 2 (September), 358–68. Rolls, Barbara J., Edward A. Rowe, and Edmund T. Rolls (1982), “How Sensory Properties of Foods Affect Human FeedingBehavior,” Physiology & Behavior, 29 (3), 409–417. Ross, Ivan (1971), “Self-Concept and Brand Preference,” Journal of Business, 44 (1), 38–50. Rucker, Derek D. and Adam D. Galinsky (2008), “Desire to Acquire: Powerlessness and Compensatory Consumption,” Journal of Consumer Research, 35 (2), 257–67. Simonson, Itamar (1990), “The Effect of Purchase Quantity and Timing on Variety-Seeking Behavior,” Journal of Marketing Research, 27 (May), 150–62. Solomon, Michael R. (1983), “The Role of Products as Social Stimuli: A Symbolic Interactionism Perspective,” Journal of Consumer Research, 10 (3), 319–29. 82 / Journal of Marketing, May 2011 Thaler, Richard (1985), “Mental Accounting and Consumer Choice,” Marketing Science, 4 (Summer), 199–214. Thomson, Matthew, Deborah J. MacInnis, and C. Whan Park (2005), “The Ties That Bind: Measuring the Strength of Consumers’ Emotional Attachments to Brands,” Journal of Consumer Psychology, 15 (1), 77–91. Thompson, Richard F. and W.A. Spencer (1966), “Habituation: A Model Phenomenon for Study of Neuronal Substrates of Behavior,” Psychological Review, 73 (1), 16–43. Tucker, William T. and John J. Painter (1961), “Personality and Product Use,” Journal of Applied Psychology, 45 (October), 325–29. Twitchell, James B. (2002), Living It Up: Our Love Affair with Luxury. New York: Columbia University Press. Veblen, Thorstein (1899), “The Preconceptions of Economic Science,” Quarterly Journal of Economics, 13 (4), 396–426. Vohs, Kathleen D., Roy F. Baumeister, Brandon J. Schmeichel, Jean M. Twenge, Noelle M. Nelson, and Dianne M. Tice (2008), “Making Choices Impairs Subsequent Self-Control: A Limited-Resource Account of Decision Making Self-Regulation, and Active Initiative,” Journal of Personality and Social Psychology, 94 (5), 883–98. Wernerfelt, Birger (1988), “Umbrella Branding as a Signal of New Product Quality: An Example of Signalling by Posting a Bond,” RAND Journal of Economics, 19 (Autumn), 458–66. Wertenbroch, Klaus and Bernd Skiera (2002), “Measuring Consumers’ Willingness to Pay at the Point of Purchase,” Journal of Marketing Research, 39 (May), 228–41. Wicklund, Robert A. and Peter M. Gollwitzer (1981), Symbolic Self-Completion. Hillsdale, NJ: Lawrence Erlbaum Associates. Copyright of Journal of Marketing is the property of American Marketing Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

2011 jm chernev et al. competing for consumer identity.pdf

0.28MB

https://smartstore.naver.com/smallme/products/5605946432

컬러링 스몰미빌딩 (20개 도시 세트) 미술교육. 창의력 발달. 해외여행. 창의력교육. 색칠공부, :