Can Brands Move In from the Outside?:How Moral Identity Enhances Out-group Brand Attitudes

1 Can Brands Move In from the Outside?: How Moral Identity Enhances Out-group Brand Attitudes WOO JIN CHOI KAREN PAGE WINTERICH* Forthcoming at Journal of Marketing *Woo Jin Choi (wjchoi@uos.ac.kr) is an Assistant Professor of Business Administration, College of Business Administration, University of Seoul, South Korea. Karen Page Winterich (kpw2@psu.edu) is an Assistant Professor of Marketing, Smeal College of Business, Pennsylvania State University, 449 Business Building, University Park, PA, 16802. This research was completed while the first author was a doctoral student at Texas A&M University with the five studies reported in this manuscript conducted in the Jeffrey S. Conant Behavioral Research Lab, Mays Business School, Texas A&M University. We appreciate the insightful comments of Lisa Bolton, Haipeng (Allan) Chen, Rajdeep Grewal, and William Ross on prior versions of this manuscript. 2 Can Brands Move In from the Outside?: How Moral Identity Enhances Out-group Brand Attitudes Consumers tend to have more favorable attitudes for their in-group brands than their out-group brands. However, little is known regarding how brand managers can overcome consumers’ negative attitudes toward out-group brands. Drawing upon the moral identity literature, the authors theorize that moral identity may enhance out-group (but not in-group) brand attitudes through decreased psychological distance. Four studies demonstrate that moral identity increases out-group brand attitudes. Two important boundary conditions of this moral identity effect are also identified. These findings suggest that brand managers seeking to overcome the less than favorable perceptions associated with out-group brands may benefit from drawing upon consumers’ moral identity in marketing communications. Theoretically, this research establishes that moral identity extends beyond prosocial behaviors to influence marketplace judgments (i.e., brand attitudes and categorization). Keywords: brand attitude, moral identity, outgroup, reference group, psychological distance 3 In one of the most beloved and long-running advertising campaigns (Adweek 2010), “Get a Mac” portrayed the Mac as the in-group computer of choice in relation to the out-group PC. Such brand positioning strategies can be effective given that brand value is enhanced for brands to which consumers relate (Chernev, Hamilton, and Gal 2011). Extensive research on reference group influence has consistently demonstrated that consumers have more favorable attitudes toward brands that are associated with the groups that they belong to and feel a part of (i.e., ingroup brands) than brands that are associated with the groups that they don’t belong to and don’t feel a part of (i.e., out-group brands) (Escalas and Bettman 2003, 2005). Conditions under which consumers’ in-group brand attitudes increase or decrease have been examined (e.g., Bearden and Etzel 1982; Chan, Berger, and Boven 2012), and research has also recognized that consumers’ negative attitudes for out-group brands may be stronger for out-groups that are avoided (i.e., dissociative; White and Dahl 2006, 2007). However, one topic that has been overlooked in the extant research, yet is important for brand managers, is an inquiry into the conditions under which consumers’ out-group brand attitudes can be improved. Starbucks was once perceived as an out-group by many given its positioning as a niche coffee brand for those with highly refined coffee tastes, yet it is now a global giant coffee chain (Daley 2009). Likewise, managers of brands that are currently perceived as out-group brands by mass market consumers are struggling to overcome the less than favorable attitudes toward them and widen their customer base to succeed (Aaker 1991; Erdem and Sun 2002). Thus, determining how consumers’ out-group brand attitudes can be improved is a critical issue for brand managers. This research seeks to provide insight into this question. Specifically, we propose that moral identity may enhance out-group brand attitudes. 4 Moral identity is the cognitive schema a person holds about the virtuosity of his or her character (Aquino et al. 2009). As detailed later, the virtuosity of one’s character refers to the accessibility of moral traits, goals, and behaviors such as honesty, compassion, or fairness (Aquino and Reed 2002; Blasi 1984). 1 Prior research shows that individuals for whom moral identity is either chronically important or temporarily accessible have more expansive group boundaries (Reed and Aquino 2003; Winterich, Mittal, and Ross 2009). That is, consumers with higher moral identity are more inclusive of outgroups such that they perceive a wider range of outgroups as their ingroups relative to someone with lower moral identity (Reed and Aquino 2003). Although the expansiveness of moral identity has been examined in regard to other individuals, we draw from the reference group literature (Escalas and Bettman 2005; White and Dahl 2007) to theorize that the expansiveness of moral identity may also influence a consumer’s brand attitude, which refers to the degree to which the consumer considers a brand as being positive or negative including quality perception (Aaker and Keller 1990). Specifically, we propose that consumers with a situationally accessible moral identity will have more positive out-group brand attitudes than those without an accessible moral identity. This effect may occur because moral identity acts as a conduit that reduces the psychological distance, which is consumers’ subjective perceptions regarding how far an object (i.e., brand) is from themselves (Trope and Liberman 2010). That is, those with an accessible moral identity will perceive less distance between themselves and brands with which their outgroups are associated, resulting in more favorable attitudes toward the brands. However, given the existing tendency to favor in-group brands (Escalas and Bettman 2005; White and Dahl 2007), moral 1 The specific traits that are most strongly associated with a moral identity differ by individual such that for some individuals generosity and compassion may come to mind whereas for others hardworking and fairness may represent moral identity. These differences are captured by each person’s cognitive schema that links various moral traits together through their associative cognitive network. See Aquino and Reed (2002) for further details. 5 identity will not influence in-group brand attitude. Thus, moral identity moderates the effect of brand group membership on brand attitude, which is mediated by psychological distance. In addition, to clarify the extent to which moral identity can enhance out-group brand attitude, we also examine the effect of moral identity on (1) aspirational versus dissociative out-group brands and (2) brands perceived as (im)moral. Our research makes several important contributions. First, we contribute to the moral identity and reference group literatures by demonstrating that moral identity can influence marketplace judgments unrelated to morality such that out-group brand attitude can be improved. Second, we demonstrate this moral identity effect occurs through decreased psychological distance. Third, we recognize the role of brand morality (Sen and Bhattacharya 2001) in the effect of consumer morality on brand attitude. In doing so, we provide practical insights for eliciting moral identity in marketing communications to enhance out-group brand attitude. We organize the paper as follows. First, we review the moral identity literature and briefly report a pilot study that demonstrates moral identity can influence marketplace judgments unrelated to morality. We then review the literature on reference group influence on brand attitude and theorize how moral identity may moderate the effect of brand group membership on brand attitude and the underlying role of psychological distance. Next, we present the results of four studies to test the moderating effect of moral identity, the underlying mechanism, and two boundary conditions. CONCEPTUAL DEVELOPMENT Moral Identity As briefly defined earlier, moral identity represents an individual’s associative cognitive network of related moral traits (e.g., being kind), feelings (e.g., concern for others), and 6 behaviors (e.g., helping others; Aquino & Reed, 2002). 2 Within this associative cognitive network, the strength of these moral associations reflects the degree to which a person’s moral identity is chronically important. For example, an individual who is characterized by high moral identity would likely have more thoughts, both in quantity and speed, regarding moral traits such as fairness and generosity as well as goals and behaviors of helping others. To be sure, this conceptualization does not indicate that an individual with low moral identity does not have any associations with moral traits or behaviors, but simply that their associative cognitive network is not as chronically strong as someone with a higher moral identity. Importantly, moral identity influences consumers’ behaviors when it is either chronically important or temporarily active (Aquino, McFerran, and Laven 2011; Reed, Aquino, and Levy 2007; Winterich, Mittal, and Ross 2009), which is consistent with effects of social identity more generally. That is, a consumer’s identity may be accessible as a result of either chronic identification or situational priming that temporarily activates an identity (Zhang and Khare 2009). Differences in these two forms of identity can lead to different outcomes (Wheeler and Berger 2007). A chronically important identity resides in one’s long-term memory whereas a temporarily accessible identity is more similar to information stored in working memory (Wheeler, DeMarree, and Petty 2008). Thus, in comparison to a chronically important identity, a temporarily accessible identity is more likely to affect one’s immediate behavior because it is 2We note that Aquino and Reed’s (2002) conceptualization of moral identity consists of two dimensions: internalization and symbolization. The former reflects the subjective experience and cognitions of moral identity that are held privately whereas the latter reflects the desire to express moral traits to the public. The internalization dimension has been referred to as moral identity centrality because the internalization of moral traits makes them easier and faster to retrieve, hence being more influential (i.e., central) on attitudes and behaviors (Aquino et al. 2009). Indeed, internalization has been a more robust predictor of moral behaviors than symbolization (Aquino and Reed 2002; Aquino et al. 2009; Reed, Aquino, and Levy 2007), particularly in regard to out-group evaluations. For instance, Reed and Aquino (2003; Study 3) show that the internalization dimension positively affects the amount of money that people are willing to donate to outgroups, whereas the symbolization dimension does not. Therefore, in this research, our theorizing regards the internalization dimension of moral identity. 7 currently more accessible than other identities. Notably, Zhang and Khare (2009) show that both chronic and temporarily accessible identities can lead to the same behavioral outcome, which is consistent with research finding that chronically important and temporarily activated moral identity have similar effects (Aquino, McFerran, and Laven 2011; Reed, Aquino, and Levy 2007; Winterich, Mittal, and Ross 2009). Based on these findings, we expect both chronic and temporarily active moral identity to result in the same effect; however, in the current research, we focus on temporarily accessible moral identity. Moral identity has emerged as a particularly useful predictor for moral behaviors such as charitable donations, altruistic helping, and on-time loan payments (Aquino et al. 2009; Hardy 2006; Herzenstein, Sonenshein and Dholakia 2011). Of particular relevance to this research is the role that moral identity plays in individuals' perceptions of outgroups, which we define as a group of individuals to which a given individual does not belong (Escalas and Bettman 2005). Research shows that moral identity mitigates negative responses to outgroups, resulting in more generous attitudes toward outgroups and greater willingness to help out-group members (Aquino and Reed 2002; Hardy et al. 2010; Reed and Aquino 2003). These effects occur due to the expanded in-group boundary associated with moral identity (Reed and Aquino 2003; Winterich, Mittal, and Ross 2009). That is, when moral identity is either chronically important or temporarily active, group boundaries tend to encompass a wider range of individuals who may otherwise be perceived as out-group members such that less distance is perceived between oneself and outgroups3 . However, to date, research has not considered whether moral identity can impact judgments unrelated to morality. 3 Trope and Liberman (2003) state that individuals perceive the distance between an object and themselves in terms of a range of measures, including time, space, and social identity. Since our focus in the current research is on 8 To the extent that research has demonstrated that consumers do not process product information in the same manner as human-related information (Yoon et al. 2006), the effect of moral identity on consumer judgments unrelated to morality is unique from prior effects of moral identity on morality-based judgments. Therefore, we conducted a pilot study to clarify if the effect of moral identity extends to judgments for which there should be no regard for morality (i.e., categorization of general objects). If moral identity influences consumers to make broader categorizations of objects (Isen and Daubman 1984; Mikulincer, Kedem, and Paz 1990; Ü lkümen, Chakravarti, and Morwitz 2010), then we can conclude that moral identity has the potential to influence marketplace judgments such as brand attitude that are unrelated to morality. In a pilot study, participants (N = 75 undergraduates; see Web Appendix for full study details) were primed so either moral identity or student identity was situationally accessible. Then they completed a categorization task adopted from Rosch (1975) and Isen and Daubman (1984) in which they indicated the extent to which each of eight items belonged to a specific category (i.e., clothing, vegetables). The results revealed that those with moral identity accessible perceived weak and moderate exemplars of the category to belong to the given category to a greater extent than those with an accessible student identity. Given that those with moral identity tend to make broader, more inclusive categorizations than those with another social identity (i.e., student identity), these results provide initial support that the expansiveness associated with moral identity impacts judgments beyond contexts associated with moral behavior. Thus, moral identity may impact brand attitude. In-group and Out-group Brand Attitudes consumers’ perceptions of distance between oneself and a brand, psychological distance refers to the consumers’ subjective perceptions of the distance between a brand and themselves. 9 Establishing that moral identity impacts general marketplace judgments of categorization, we consider how moral identity may differentially influence attitudes toward in-group and outgroup brands. Research has established that people tend to prefer things that are similar to themselves to boost their self-esteem and self-evaluations (Brewer 1991). That is, people categorize themselves in terms of groups of individuals to which they are similar or different to determine their social identities (Tajfel and Turner 1979; Turner 1985). Based on these categorizations, people respond more positively to their ingroup, which is defined as a group of individuals to which one feels they belong, than their outgroup, defined earlier as a group of individuals to which one does not feel they belong (Escalas and Bettman 2005). These in-group preferences extend to consumer behaviors and marketplace judgments (Berger and Heath 2007; Berger and Rand 2008; Escalas and Bettman 2005; White and Dahl 2006, 2007). For example, Escalas and Bettman (2005) demonstrated that consumers connect more strongly (i.e., self-brand connection) to brands associated with their ingroups than their outgroups due to a greater perceived “fit” between themselves and the in-group brands. Given these in-group preferences, prior research on reference groups has generally been conducted along two lines. First, boundary conditions of in-group preferences have been identified (e.g., public vs. private consumption and luxury vs. necessity: Bearden and Etzel 1982; need for uniqueness: Chan, Berger, and Boven 2012; self construal: Escalas and Bettman 2005; compromised individuality by mass production: White and Argo 2011). Second, White and Dahl (2006, 2007) demonstrate that the type of outgroup may impact consumer’s response. Yet, research examining the conditions under which consumers’ less than favorable attitudes toward out-group brands can be improved is limited. This question is the focus of the current research for which we consider the moderating role of moral identity. 10 The Moderating Effect of Moral Identity on In-group and Out-group Brand Attitudes How might consumers’ moral identity influence the less than favorable attitudes toward out-group brands? First, while moral identity influences one’s regard for out-group members (Reed and Aquino 2003), the effect of moral identity on moral regard for out-groups may not occur for brand attitude due to the lack of relevance to morality and the different processes used for brand judgments (Yoon et al. 2006). On the other hand, as in human relationships, consumers develop relationships with brands (Fournier 1998). More specifically, the in-group preference initially established among groups of individuals exists for brand attitudes (Escalas and Bettman 2003, 2005; White and Dahl 2007). Thus, it is plausible that the expansiveness of moral identity that improves moral regard for out-groups (Reed and Aquino 2003; Winterich, Mittal, and Ross 2009) may also overcome unfavorable out-group brand attitudes. Moreover, our pilot study demonstrated that the expansiveness associated with moral identity can impact judgments unrelated to morality such that moral identity may carryover to impact brand attitude. Thus, we expect moral identity to moderate attitudes toward in-group and out-group brands. The distinction between ingroups and outgroups can vary according to context (Berger and Heath 2007; Brewer and Gardner 1996) since the preference for ingroups occurs from the association with group members (Bearden and Etzel 1982). When these group associations are altered, then consumers’ brand attitudes may also be altered. Gómez and colleagues (2011) find that as individuals categorize an outgroup in a more inclusive manner, they respond more favorably to this group because this inclusive categorization provides a common ground for the two groups. Similarly, thinking about the shared characteristics of women and men can reduce negative gender stereotypes (i.e., math performance) because the focus on similarities reduces the focus on differences between genders (Rosenthal and Crisp 2006). 11 Drawing from these findings, we theorize that moral identity, which increases perceived similarities with out-group members and expands the psychological boundaries of group membership (Reed and Aquino 2003; Winterich, Mittal, and Ross 2009), will affect consumers’ brand attitudes. Specifically, we predict that moral identity will impact the degree to which a consumer perceives a brand as being an out-group brand. That is, consumers with accessible moral identity will perceive less distance between themselves and out-group brands, thereby, responding more favorably toward these brands relative to those without accessible moral identity. To be clear, we do not propose that the differences between consumers and the individuals they associate with an out-group brand are physically decreased; rather, consistent with prior research (Gómez et al. 2011; Rosenthal and Crisp 2006), consumers may simply focus on similarities, thereby perceiving less distance and being more inclusive. If moral identity improves attitudes toward out-group brands, how will it impact attitudes toward in-group brands? Consumers tend to naturally perceive more similarities and feel closer to in-group brands (Escalas and Bettman 2003, 2005), evaluating them favorably (White and Dahl 2007). Therefore, we posit that consumers will have more positive attitudes toward ingroup brands regardless of moral identity because their boundaries already extend to ingroups, unlike outgroups, and thus are not subject to the expanding group boundaries of moral identity. This effect is theorized as follows: H1: Moral identity will moderate the effect of brand group membership on brand attitude such that moral identity will enhance out-group brand attitudes but will not influence in-group brand attitudes. The Role of Out-group Type Thus far, our theorizing has utilized the dichotomous distinction of ingroups and outgroups, following previous research (Berger and Heath 2007; Escalas and Bettman 2005; 12 White and Argo 2011). However, outgroups can be further classified as dissociative groups and aspirational groups. A dissociative group refers to a group for which individuals hold negative associations and wish to avoid (White and Dahl 2007; McFerran et al. 2010), whereas an aspirational group refers to a group for which an individual has positive associations and desires to belong (Escalas and Bettman 2003). Given these positive versus negative associations, aspirational groups do not have the same effects as dissociative groups (McFerran et al. 2010). Specifically, consumers respond most negatively to dissociative groups (White and Dahl 2007), whereas they tend to favor brands associated with aspirational groups (Escalas and Bettman 2003). Although dissociative and aspirational groups are similar in that consumers do not belong to either group, the negative out-group response is much stronger for dissociative groups. H1 theorized that moral identity could improve out-group brand attitudes, without specifying effects for dissociative versus aspirational groups. Given that consumers respond favorably to aspirational group brands similar to in-group brands (Escalas and Bettman 2003) but negatively toward dissociative group brands (McFerran et al. 2010; White and Dahl 2007), we theorize that moral identity will only affect attitudes toward dissociative group brands (vs. ingroup and aspirational group brands). Specifically, since consumers tend to avoid dissociative groups, the expansiveness of moral identity will be necessary to reduce the psychological distance to the dissociative group brand, resulting in more favorable dissociative group brand attitudes. If moral identity enhances dissociative group brand attitudes, this would suggest that the effect of moral identity is not limited to neutral outgroups but extends to outgroups that have stronger negative associations. In contrast, if a brand is associated with an aspirational outgroup, consumers favor the brand due to their desire to associate with this aspirational group. In this case, the expansiveness of moral identity does not influence brand attitude since the 13 psychological distance to the brand, similar to in-group brands, is already minimal. Thus, we expect moral identity to only influence dissociative group brand attitudes. Therefore, H2: The moderating effect of consumers’ moral identity on the effect of brand group membership on brand attitude will be influenced by the type of outgroup associated with the brand. Specifically, moral identity will enhance brand attitudes for brands associated with dissociative groups, but this effect will be weaker for brands associated with aspirational groups, similar to the case for ingroup brand attitudes. The Mediating Role of Psychological Distance Based on our theorizing that moral identity enhances out-group (i.e., neutral and dissociative) brand attitudes because moral identity decreases the distance perceived between oneself and out-group brands, we predict that the moral identity effect will be mediated by the psychological distance between oneself and a brand. The social psychology literature demonstrates that psychological distance influences the use of conflict between groups as a basis for evaluation. Specifically, people are more likely to treat others favorably if they perceive that others are close to them (Liberman, Samuels, and Ross 2004). For example, undergraduate students allocated more money to those listed higher on their list of the 100 closest people to them (Jones and Rachlin 2006) such that psychological distance influenced altruism. Moreover, individuals who categorize out-group members using a common identity with them and thus perceive a greater connection to them are less willing to display bias toward their own group (Gόmez and colleagues 2011). In a similar vein, Winterich, Mittal, and Ross (2009) find that the psychological distance between consumers and out-group members mediates their willingness to donate money to the outgroup. Though psychological distance tends to be examined as the distance between two individuals rather than between a consumer and a brand, it is possible that the perception of 14 distance from others applies to brands given the relationships and group associations with brands (Escalas and Bettman 2003, 2005; Fournier 1998). Thus, we theorize that moral identity will decrease the psychological distance between a consumer and their out-group brands, resulting in more favorable out-group brand attitudes. That is, the association of human outgroups with brands will result in an effect of moral identity on the psychological distance between the consumer and the brand. In other words, the expansiveness associated with moral identity that decreases psychological distance should carryover to judgments regardless of the judgment type. Formally, H3: The moderating effect of consumers’ moral identity on the effect of brand group membership on brand attitude will be mediated by the psychological distance between the consumer and the brand. Next, we test our hypotheses in a series of four studies. Study 1 tests the moderating effect of moral identity on brand attitudes toward self-identified in-group and out-group brands (H1), using temporarily accessible moral identity versus student identity. Study 2 replicates the findings of Study 1 for dissociative out-group brand attitudes but not for aspirational out-group or in-group brand attitudes (H2). Study 3 provides managerial implications by examining the effect of temporarily activating a moral identity versus a business identity in an advertisement for a hypothetical brand. In doing so, we examine the mediating role of the psychological distance between the consumer and a brand, supporting H3. Study 4 replicates the mediating role of psychological distance on out-group brand attitude while taking into account a brand’s morality in addition to consumers’ moral identity. STUDY 1 Study 1 was designed to examine H1, which proposed that MI would enhance out-group brand attitude but not in-group brand attitude, such that in-group brand attitude would remain 15 constant irrespective of MI. We tested this hypothesis by using self-identified in- and out-group brands and temporarily activating either a moral (MI) or student (SI) identity. Method Participants and Procedure. Ninety undergraduates (57% male) participated in this study in exchange for course credit. A 2 (brand group membership: ingroup vs. outgroup) X 2 (accessible identity: SI vs. MI) mixed design with brand group membership as a within-subjects factor was used. The experiment consisted of three parts; each presented as an unrelated study with a distinct cover story. First, following Escalas and Bettman (2005), participants were asked to list one group that they belonged to and felt a part of and another group that they didn’t belong to and didn’t feel a part of. Immediately after listing each group, they self-identified one brand associated with their stated ingroup and one brand associated with their outgroup. For example, the following groups (brands) were listed: Zeta Tau Alpha (Columbia), Sigma Phi Epsilon (Polo), and Kappa Delta (Vera Bradley) for ingroups; Athletes (Adidas), Korean students (Hello Kitty), and Greeks (Sperry) for outgroups. Note that the groups and brands were idiosyncratic, consistent with White and Dahl (2007). Next, each participant was primed with one of two randomly assigned social identities: SI or MI. Then, participants indicated their attitudes toward the two brands that they identified in part 1 along with two filler brands. Brand attitudes were assessed using four items (in-group: α = .92, out-group brand: α =.95) on a 7-point scale (Sood and Keller 2012; White and Dahl 2007). At the end of the study, participants were asked to guess the purpose of this study, but no one guessed it correctly in this or subsequent studies. All measures used in this and all subsequent studies are summarized in Appendix A. 16 Social Identity Prime. In the “writing assessment” task, participants were asked to write three short essays about themselves that included a set of nine words, corresponding to the identity condition. The two sets of nine words utilized were: book, car, chair, computer, desk, pen, street, table and trash can (SI) and caring, compassionate, fair, friendly, generous, helpful, hardworking, honest, and kind (MI; Reed, Aquino, and Levy 2007). Using the nine words corresponding to their condition, participants were asked to write about: 1) several things they experienced and enjoyed daily as a student (moral person), 2) a few reasons why being a student (moral person) is important to them, and 3) one of the most positive experiences that they had as a student (moral person; adapted from LeBoeuf, Shafir, and Bayuk 2010). Results and Discussion Manipulation Check. An ANOVA tested the main effect of identity condition (SI vs. MI) with each of the four manipulation check items as dependent variables. MI (vs. SI) participants indicated their writing was more reflective of them as a moral person (MMI = 5.93 vs. MSI = 5.04; F(1,88) = 9.89, p < .01). Those in the SI (vs. MI) condition indicated their writing was more reflective of them as a student (MMI = 5.14 vs. MSI = 5.68; F(1,88) = 2.96, p = .08). The other two items did not differ by identity condition (p > .10). Brand Attitude. We conducted a repeated-measures ANOVA with brand group membership, accessible identity, and their interaction with brand attitude as the dependent variable. Brand group membership (ingroup = 0, outgroup = 1) was a within-subjects factor and identity condition (SI condition = 0, MI condition = 1) was a between-subjects factor. Brand group membership was significant (F(1,88) = 100.87, p < .01) such that out-group brand attitudes were less positive (M = 4.55) than in-group brand attitudes (M = 6.51). Importantly, the 17 interaction of identity condition and brand group membership was significant (F(1,88) = 11.96, p < .01). Further analyses revealed that the effect of identity condition on in-group brand attitudes was not significant (F(1,88) = 1.71, p = .19) but was significant for out-group brand attitudes (F(1,88) = 8.89, p < .01). Those in the MI condition had more favorable attitudes toward outgroup brands than those in the SI condition (MMI = 5.14 vs. MSI = 4.03), supporting H1 (see Figure 1). There were no effects on the attitude toward filler brands (p > .50). The effect size for the interaction was small: =.015 and Cohen’s d = .24, which is likely due to un-captured variation arising from self-identified in-group and out-group brands. [Insert Figure 1 about here] Discussion. Study 1 demonstrates that consumers with MI have more favorable attitudes toward their out-group brands than those with SI. However, attitudes toward in-group brands were equally favorable, regardless of identity condition. Thus, Study 1 provides support for H1 such that consumers’ MI can enhance attitudes toward out-group brands. Next, we examine whether MI enhances out-group brand attitudes even when brands are associated with dissociative outgroups versus aspirational outgroups and ingroups, testing H2. STUDY 2 Participants and Procedure. Eighty-four undergraduate students (53% male) participated in this study in exchange for course credit. A 3 (brand group membership: in-group vs. dissociative out-group vs. aspirational out-group) X 2 (accessible identity: SI vs. MI) mixed design with brand group membership as a within-subjects factor was used. As in Study 1, the experiment consisted of three parts each with a distinct cover story. Participants first listed one group and a corresponding brand for each of the following groups: ingroups, aspirational out- 18 groups, and dissociative out-groups. Definitions for each group were provided to participants to aid in accurate self-identification of groups (Escalas and Bettman 2003, 2005; White and Dahl 2007). Examples of groups (corresponding brands) included: Business Student Council (Apple), AITP (Association of Information Technology Professionals; HP), and Pi Beta Phi (Nike) for ingroups; Honors Program (Tory Burch), Lions Club (Juicy Couture), and Phi Beta Lambda (Vineyard Vines) for aspirational groups; and Greek Life (Sperry), Sorority (Polo Ralph Lauren), and student government (American Eagle) for dissociative groups. As in Study 1, the groups and brands were idiosyncratic. In part 2, participants completed a randomly assigned (SI vs. MI) essay task used in Study 1, which was followed by the same manipulation check items. In part 3, the three brands identified in part 1 were presented in random order and evaluated using the same four items as Study 1 (dissociative group brand: α = .93, aspirational group brand: α = .93, ingroup brand: α = .91). Results and Discussion Manipulation Check. MI (vs. SI) participants indicated their writing was more reflective of them as a moral person (MMI = 5.44 vs. MSI = 4.73; F(1,82) = 4.84, p < .05). Those in the SI (vs. MI) condition indicated their writing was more reflective of them as a student (MMI = 4.88 vs. MSI = 5.64; F(1,82) = 5.23, p < .05). The other two items did not differ by identity condition (p > .80). Brand Attitude. A repeated-measures ANOVA with brand group membership as a within-subjects factor, identity condition as a between-subjects factor, and their interaction indicated that the main effect of brand group membership on brand attitude was significant (F(2,164) = 209.32, p < .01; MIn-group = 6.18, MAspirational-group = 6.10, MDissociative-group = 2.78). More importantly, the interaction was significant (F(2,164) = 3.73, p < .05). As depicted in Figure 2, 19 planned contrasts demonstrated that the effect of MI was significant only for the dissociative group brands such that dissociative group brand attitudes were more positive in the MI condition than those in the SI condition (MMI = 3.14 vs. MSI = 2.38; F(1,164) = 4.33, p < .05). This was not the case for aspirational group brands (F(1,164) = .38, p > .53) nor in-group brands (F(1,164) = .90, p > .34). The effect size of the interaction was small-to-medium ( =.044; Cohen’s d = .42), which, as in Study 1, may be due to variation from self-identified brands. [Insert Figure 2 about here] Discussion. This study clarifies the moderating effect of MI on out-group brand attitude: MI enhances brand attitudes even for dissociative group brands which are avoided, but not for the desired aspirational group brands, supporting H2. One question that may arise is the distinction between these dissociative group brands and the neutral out-group brands examined in Study 1. Comparing the magnitude of brand attitudes suggests that the outgroups and corresponding brands identified in Study 1 (M = 4.58; also Study 3: M = 4.84) were not disliked and avoided as strongly as were those listed for the dissociative groups in Study 2 (M = 2.76). This pattern is consistent with distinctions between neutral and dissociative outgroups (White and Dahl 2007), but also makes it more notable that MI enhanced dissociative group brand attitudes. These findings extend White and Dahl (2007)’s research by identifying additional conditions (i.e., MI) under which dissociative group brand attitudes can be improved. Importantly, the attenuation of the MI effect for aspirational group brands was due to the favorable attitudes toward aspirational group brands regardless of MI rather than an inability of MI to enhance unfavorable brand attitudes. Recognizing that the effect of MI extends to dissociative group brands, we next seek to examine the underlying process of the effect of MI on out-group (vs. in-group) brand attitudes, 20 testing H3. In doing so, we also examine whether consumers’ MI can be activated through an advertisement. Though previous studies on MI (Aquino, McFerran, and Laven 2011; Reed, Aquino, and Levy 2007; Winterich, Mittal, and Ross 2009) have temporarily activated MI, to the best of our knowledge, this study is the first attempt to temporarily activate MI via an advertisement. Demonstrating the ability to prime this distinct social identity in an advertisement would offer managers the opportunity to temporarily activate moral identity in their marketing communications to improve brand attitude. In addition, we use a hypothetical brand in the next two studies to overcome potential confounding effects arising from individual differences in prior knowledge and experience when examining self-identified brands, used in Studies 1 and 2. STUDY 3 The objective of Study 3 was two-fold. First, we tested H3, which theorizes that the effect of MI is mediated by psychological distance such that MI results in less psychological distance to out-group brands. Second, we primed one of two social identities in an advertisement for a hypothetical brand, thereby making the identity temporarily accessible. The use of a hypothetical brand allowed the brand to be described as either an in- or out-group brand. Since brand associations and attachment require time to develop through repeated interactions (Hazan and Shaver 1994; Swaminathan, Stilley, and Ahluwalia 2009), a hypothetical brand will likely not result in the same strength of in-group preference observed in past studies, but the hypothesized effect should emerge regardless. That is, participants with MI accessible will have more favorable attitudes toward out-group brands compared to those with another accessible social identity, but the social identity primed in the advertisement should not influence in-group brand attitudes, consistent with H1. Method 21 Stimuli Development. We used the hypothetical watch brand, Mardi, which was previously used as a portable clock brand by Swaminathan, Stilley, and Ahluwalia (2009). For the accessible identities, we selected MI and business identity (BI) because BI was relevant to our sample of undergraduate business students as well as for advertising watches. To prime MI and BI, two distinct advertisements activating the corresponding identity were developed (see Appendix B for the advertisements). Two short pretests verified: 1) the advertisements temporarily activated the corresponding identity as intended without altering advertisement evaluations and 2) perceptions of brand morality4 were not influenced by the identity prime. In the first pretest (N = 67 undergraduates), participants viewed either the MI or BI advertisement and evaluated the advertisement with two items (r = .65, p < .01; see Appendix A). An ANOVA revealed that the effect of identity was not significant (F(1,65) = 2.17, p > .10), indicating that attitudes toward the advertisement did not vary by identity prime. Thus, any differences in brand attitudes toward the Mardi brand assessed in the main study are not attributable to differences in advertisement evaluations from the identity prime. In addition, we measured the manipulation check items from prior studies. Conducting an ANOVA, we find that MI (vs. BI) condition participants thought more about themselves as moral individuals (MMI = 4.37 vs. MBI = 2.97; F(1,65) = 12.02, p < .001) whereas BI (vs. MI) condition participants thought more about themselves as business leaders (MMI = 3.69 vs. MBI = 4.63; F(1,65) = 5.18, p = .02). The other two items did not differ by identity condition (p > .30). Thus, the advertisement successfully primed BI or MI without altering advertisement evaluations. 4 We define brand morality as the degree to which a consumer perceives a brand to be virtuous. Similar to how a consumer’s moral identity consists of moral characteristics (Aquino and Reed 2002), a brand’s morality may be based on perceptions of the brand’s virtuosity or moral characteristics, such as the degree to which the brand expresses moral sentiments and is involved in moral behaviors. 22 In the second pretest (N = 80 undergraduates), participants were randomly assigned to view the advertisement associated with one of the two identity conditions (MI vs. BI). They indicated their attitude toward the advertisement as in the first pretest (r = .91, p < .01) as well as the quality and morality of the Mardi brand (see Appendix A). An ANOVA revealed that the identity condition did not influence advertisement evaluations (F(1,78) = 2.13, p = .14), the perceived quality of the Mardi brand (F(1,78) = 2.76, p > .10; MMI = 4.64, MBI = 5.18), or the perceived morality of the Mardi brand (F(1,78) = .11, p > .73; MMI = 4.45, MBI = 4.55). This latter result, though perhaps surprising, is particularly important as it demonstrates that eliciting MI in an advertisement does not significantly alter the perceived brand morality. Thus, the advertisement activated MI but did not influence consumers’ perceptions of brand quality, brand morality, or advertisement evaluations so we used these two advertisements in the main study. Participants and Procedure. A total of 109 undergraduates (39% male) participated in exchange for course credit. Participants were randomly assigned to one of four conditions in a 2 (brand group membership: ingroup vs. outgroup) X 2 (accessible identity: BI vs. MI) betweensubjects design. The study consisted of two parts that were presented as unrelated studies with separate introduction pages. In part 1, participants listed one group which they perceived as an ingroup and one as an outgroup. In part 2, a hypothetical watch brand, Mardi, was presented and participants were informed that they saw several members of their self-identified ingroup (outgroup) wearing Mardi watches, depending on their randomly assigned brand condition. Then, participants were shown either the BI or MI advertisement, depending on their randomly assigned identity condition. They evaluated the advertisement with the two items used in the 23 pretests (r = .89, p < .01)5 and then indicated their attitudes toward the Mardi brand with the four items used in Studies 1 and 2 (α = .91 for both in-group and out-group brand attitudes). Last, participants completed the Inclusion of Others in Self scale adapted for the brand (IOS; Aron, Aron, and Smollan 1992) to measure the psychological distance between themselves and the brand (see Appendix A). This measure assesses perceptions of interconnectedness between oneself and another person and was adapted to assess perceptions of interconnectedness between oneself and a brand such that this measure taps into consumers’ general perceptions of the distance between themselves and a brand. Higher scores indicate less psychological distance. Results and Discussion Brand Attitude. A 2 (accessible identity) X 2 (brand group membership) ANOVA on brand attitude revealed a significant interaction effect for identity and brand group membership (F(3,105) = 5.61, p = .01). No other effects were significant (p’s > .40). As expected, planned contrasts demonstrated that when Mardi was presented as an out-group brand, those in the MI condition evaluated it more favorably than those in the BI condition (MMI = 5.18, MBI = 4.51, F(1,105) = 4.23, p < .05). In contrast, when the brand was presented as an in-group brand, the difference between the MI and BI condition was not significant (MMI = 4.80, MBI = 5.25, F(1,105) = 1.71, p = .19; see Figure 3). Thus, activating consumers’ MI as opposed to a BI in an advertisement enhances brand attitudes when the brand is affiliated with an outgroup, but has no effect when the brand is affiliated with an ingroup, supporting H1. We did not find the main effect of brand group membership evident in Studies 1 and 2, but we believe this is due to the use of a hypothetical brand for which participants have not developed associations through repeated interactions over time (Hazan and Shaver 1994). Additionally, the effect size of the interaction 5 As in the pretests, advertisement evaluations were not influenced by identity condition, brand group membership condition, or their interaction (p’s > .10) and are not discussed further. 24 was moderate ( = .077; Cohen’s d = .57). This is likely larger than that of Studies 1 and 2 due to the reduced variation in participant responses when using one hypothetical brand rather than self-identified real-world brands. [Insert Table 1 and Figure 3 about here] Mediating Role of Psychological Distance. To test H3 regarding the mediating role of psychological distance, we used a bootstrapping approach for assessing mediated moderation (Preacher, Rucker, and Hayes 2007). Recall that psychological distance was measured by IOS (see Table 1 for the IOS means). First, we found that the 95% CI for the higher order identity prime × brand group membership interaction excluded zero (.0063 to .8029). Thus, we proceeded to examine the pattern of mediation by in-group and out-group brands. The 95% CI for the conditional indirect effect of identity did not include zero for the out-group brand (.0548 to .6363), but it did include zero for the in-group brand (-.3151 to .2228). These findings support H3 such that psychological distance mediated the effect of identity for out-group brands. 6 We do note that it may be surprising that the IOS means for the ingroup are not higher, but we believe these relatively low IOS in-group means be due to the hypothetical brand used in this study for which the main effect of brand group membership found in Studies 1 and 2 was not replicated. Discussion. In Study 3, we replicated the moderating effect of MI on brand group membership on brand attitude when MI was activated through an advertisement. This method provides practical implications for marketing managers to utilize MI in their marketing 6 We utilized Model 2 of Preacher, Rucker, and Hayes (2007), which examines mediation when the effect of brand group membership on the mediator is moderated by moral identity. Support for this model confirms our theorizing that brand group membership interacts with moral identity to influence the psychological distance between a consumer and a brand, which in turn affects brand attitude. We also employed Model 3 to examine the alternative possibility that the moderation was occurring not before the mediator but rather after the mediator such that brand group membership predicted psychological distance and identity only moderated the effect of distance on brand attitude. For this model, the 95% CI for the indirect effect of identity included zero for both the in-group brand (-.0473 to .4155) and out-group brand (-.0432 to .3922), indicating that the alternative model is not supported. The same analysis was conducted for Study 4 with support for Model 2. 25 communications. Additionally, we demonstrated the mediating role of psychological distance for the effect of identity on out-group brands, thereby providing insight into the process through which MI enhances out-group brand attitudes. These findings are particularly noteworthy in that we used a hypothetical brand, thus avoiding potential confounds from using real brands. One possible concern is that the moral words used in the MI advertisement influenced perceptions of brand morality, thereby enhancing attitudes toward out-group brands, but our pretest indicated this did not occur. Furthermore, the analysis in the main study demonstrated that those in the MI condition had more favorable out-group brand attitudes due to decreased psychological distance between themselves and the out-group brand. If the more favorable brand attitudes in the MI condition resulted from perceptions of brand morality, then psychological distance should not be influenced and brand attitudes should be more favorable in the MI condition than the BI condition regardless of brand group membership. This was not the case. Thus, the MI advertisement enhanced out-group brand attitude relative to another social identity (i.e., BI) due to the decrease in psychological distance to the out-group brand rather than through brand morality. However, we next consider the role of brand morality. STUDY 4 Though our focus has been on consumers’ MI thus far, the role of brand morality such as CSR (corporate social responsibility) behaviors on consumer attitudes is an emerging area of research. For example, brands’ CSR activities can positively impact brand attitudes (Brown and Dacin 1997), purchase intentions (Sen and Bhattacharya 2001), and donations to corporatesupported nonprofit organizations (Lichtenstein, Drumwright and Braig 2004). This positive effect of CSR may occur by strengthening consumers’ identification with the brand (Sen and Bhattacharya 2001), which is consistent with decreased psychological distance. Additionally, 26 Verrochi, Reed, and Tong (2009) find that MI predicts attitudes toward firms engaging in CSR activities through values-driven attributions of the firm. However, this stream of research has focused on either (1) brands’ morality (i.e., CSR activities) without considering consumers’ morality or (2) consumers’ morality without considering varying levels of brand morality. Therefore, in our last study, we take into account both brands’ and consumers’ morality. We propose that the effect of brand morality on brand attitude will be jointly influenced by consumers’ MI and brand group membership given that CSR effects may occur through one’s connection to (i.e., psychological distance from) a brand. We theorize that when a brand is perceived as being moral (immoral), consumers will have more (less) favorable brand attitudes, regardless of MI and brand group membership. This effect should occur because consumers, even those without an activated MI, have a base level of moral standards (Bandura 1999) and should respond (un)favorably to firms (not) engaging in CSR activities, consistent with past research (Lichtenstein, Drumwright and Braig 2004). That is, when a brand is perceived as clearly (im)moral, consumers will (not) be closer to the brand and (dis)like the brand. In contrast, when a brand is not clearly moral but reveals a tendency or possibility of being moral through CSR activities, MI may moderate the effect of brand group membership on brand attitude, as proposed in H1. That is, if a brand’s morality is unknown, the expansiveness of MI can be exercised such that activated MI will enhance out-group brand attitude. This effect should occur through decreasing psychological distance to the out-group brand. This theorizing manifests in a three-way interaction of brand morality, consumer MI, and brand group membership on brand attitude. Specifically, we hypothesize: H4: The moderating effect of consumers’ moral identity on the effect of brand group membership on brand attitude will be influenced by brand morality. When a brand is perceived as being moral or immoral, moral identity and brand group 27 membership will not influence brand attitude. In contrast, when brand morality is neutral, moral identity will enhance out-group brand attitude. We tested H4 in this study. Additionally, as in Study 3, we examined support for H3. Method Stimuli Development. We used the hypothetical watch brand name, Mardi, as in Study 3. Brand morality was manipulated by having participants read a newspaper article describing Mardi’s decision to contribute/not contribute/undecided to an educational program to reduce the educational gap between the rich and the poor in the U.S. for the moral/immoral/neutral conditions, respectively (see Appendix C). Pretest participants (N=55 undergraduates) were randomly assigned to one of the three brand morality conditions and read the corresponding newspaper article. Then, they indicated perceived brand morality on five items (α = .90; see Appendix A). An ANOVA confirmed a significant effect of brand morality condition on brand morality perceptions (F(2,51) = 7.27, p < .01). Additionally, one-sample t-tests were conducted for the moral and immoral conditions against the mean of the neutral condition (4.14). In the moral (immoral) condition, the mean was significantly higher (lower) than the neutral condition (Mmoral = 4.82, t(17) = 2.27, p < .05; Mimmoral = 3.44, t(17) = -3.54, p < .01). Article realism and relevance were also examined, but brand morality did not affect either (ps > .70; see Appendix A). Taken together, these results demonstrate that the brand morality primes worked as intended. Participants and Procedure. Participants were 317 undergraduates (49% male) who participated for course credit. A 2 (brand group membership: ingroup vs. outgroup) X 2 (accessible identity: SI vs. MI) X 3 (brand morality: moral, immoral, and neutral) betweensubjects design was used. The experiment was presented as three separate studies. Part 1 was identical to Study 3 in which participants listed one ingroup and one outgroup. In part 2, they 28 completed either the MI or SI writing assessment task as in Studies 1 and 2. In part 3, they were randomly assigned to one of six conditions: moral/in-group brand, neutral/in-group brand, immoral/in-group brand, moral/out-group brand, neutral/out-group brand, and immoral/outgroup brand. The pretested newspaper article about the hypothetical watch brand, Mardi, was presented along with the information used in Study 3 that several members of their self-identified ingroup (outgroup) wear Mardi watches. Then, participants indicated their attitude toward the Mardi brand with the same four items used in earlier studies (α = .88). Results and Discussion Manipulation Check. As expected, MI (vs. SI) participants indicated their writing was more reflective of them as a moral person (MMI = 5.80 vs. MSI = 4.50; F(1,315) = 46.06, p < .01). Those in the SI (vs. MI) condition indicated their writing was more reflective of them as a student (MMI = 4.84 vs. MSI = 5.73; F(1,315) = 21.07, p < .01). Brand Attitude. To examine the three-way interaction proposed in H4, we conducted an ANOVA on brand attitude with identity condition, brand morality, brand group membership and the interaction terms of these three variables. The effect of brand morality was significant (F(2,305) = 38.93, p < .01): the moral brand was perceived more favorably than the neutral and immoral brand and the neutral brand was favored over the immoral brand (MMoral = 4.81 > MNeutral = 4.49 > MImmoral = 3.47). The effect of brand group membership was also significant (F(1,305) = 3.81, p = .05; MIN = 4.38 > MOUT = 4.13) such that consumers favored in-group brands more than out-group brands. Importantly, the three-way interaction of identity condition, brand morality, and brand group membership was significant (F(11,305) = 4.15, p = .01). Thus, we next examined the effect of identity prime by each brand morality level. 29 First, the two-way interaction between identity condition and brand group membership was not significant for the moral brand (F(1,102) = .84, p > .36) or the immoral brand (F(3,98) = .65, p > .42; see Figure 4). In contrast, the two-way interaction between identity condition and brand group membership was significant for the morally neutral brand (F(3,105) = 13.10, p < .001). Planned contrasts revealed that when Mardi was presented as an out-group brand that was morally neutral, MI (vs. SI) resulted in more favorable brand attitudes (F(1,105) = 13.37, p < .001; MMI = 4.97, MSI = 3.96). When Mardi was presented as an in-group brand, there was no effect of identity (F(1,105) = 2.14, p > .10; MMI = 4.71, MSI = 4.31). The effect size of the interaction of identity condition and brand group membership in the neutral condition was large (Cohen 1992; =.110; Cohen’s d = .70). These results support H4. [Insert Figure 4 about here] Mediating Role of Psychological Distance. We used the bootstrapping approach from Study 3 to assess the mediating role of psychological distance in the morally neutral brand condition (see Table 1 for the IOS means). First, the 95% CI for the accessible identity × brand group membership interaction excluded zero (.0113 to .7175), so we examined the pattern of mediation for in-group and out-group brands. The 95% CI for the conditional indirect effect of identity condition did not include zero for the out-group brand (.3632 to 1.2047) but included zero for the in-group brand (-.2518 to .5262). These findings indicate that psychological distance mediated the effect of identity condition for the out-group brand, supporting H3 and consistent with Study 3. As anticipated, psychological distance did not underlie attitudes toward moral and immoral brands since there was no effect of identity. Discussion. Study 4 identified a boundary condition of the MI effect. When brands are perceived as moral or immoral, consumers’ MI does not appear to influence the effect of brand 30 group membership on brand attitude since brand (im)morality perceptions tend to override ingroup preferences that otherwise occur. However, when brand morality is neutral, MI can enhance out-group brand attitude such that morally neutral out-group brands may benefit more from MI than moral or immoral brands. We also replicated the mediating role of psychological distance found in Study 3. Together, these four studies provide strong support for our theorizing that MI can aid in overcoming the negative attitudes otherwise associated with out-group brands. GENERAL DISCUSSION In a series of four studies, this research offers important insights into the effects of moral identity in the marketplace beyond consumer donations and prosocial behavior. A pilot study provides initial evidence that the effect of moral identity extends beyond moral behaviors to marketplace judgments unrelated to morality (i.e., object categorizations). Then, theorizing how moral identity may influence attitudes toward in-group and out-group brands, we demonstrate that moral identity can increase out-group brand attitude (Study 1). Importantly this occurs not only for neutral out-group brands, but also for dissociative out-group brands, which tend to have stronger negative associations (Study 2). Additionally, we show that while consumers’ moral identity does not influence out-group brand attitude for moral or immoral brands, moral identity is beneficial to out-group brands with more neutral brand morality perceptions (Study 4). Moreover, this effect occurs because consumers with an accessible moral identity perceive less distance between themselves and out-group brands (Studies 3 and 4). In the case of in-group brands as well as aspirational group brands, since the perceived distance between consumers and these brands is minimal, moral identity does not exert an influence. Hence, MI influences outgroup brand attitude through reduced psychological distance due to smaller distinctions between objects rather than a decrease in prejudice toward outgroups. 31 Theoretical Contributions for the Role of Moral Identity in Marketplace Judgments Our primary theoretical contributions are to literature on reference groups, moral identity, and accessible and chronic identities. First, we contribute to the reference group literature by identifying one of the few ways in which unfavorable attitudes toward outgroups can be overcome: moral identity. While the extant reference group literature (e.g., Chan, Berger, and Boven 2011; White and Argo 2011) has identified the unfavorable attitudes toward outgroups, we provide new insights into the process through which this effect occurs by examining psychological distance. To be sure, past work has shown that the effect of moral identity on charitable donations to ingroups and outgroups was mediated by IOS (Winterich, Mittal, and Ross 2009). However, the present research moves beyond this existing research by demonstrating that the psychological distance between a consumer and a brand can be assessed. Moreover, psychological distance constitutes the underlying process affecting brand attitude rather than being limited solely to processes involving judgments of other individuals. We also augment the moral identity literature by demonstrating that the effect of moral identity extends beyond morally relevant judgments and behaviors (Aquino, McFerran, and Laven 2011; Reed, Aquino, and Levy 2007; Skarlicki, van Jaarsveld, and Walker 2008) to brand attitudes and categorization, substantially extending the scope of moral identity. We believe these results will open up future research to consider the unique benefits of moral identity in marketing strategy that extend beyond that of brand morality and may be distinct from the effects of other social identities used in marketing communications. Moreover, our results provide insight into the relationship between accessible and chronic identities. In the present studies, the strength of the accessible moral identity was positively correlated with chronic moral identity (r 32 = .17 ~ .25, ps < .05) 7 , which is consistent with findings that the effect of a temporarily accessible identity is stronger when it is compatible with an individual’s chronic identity (Zhang and Khare 2009). Furthermore, in additional studies not reported here, chronic moral identity increased out-group brand attitude, demonstrating that chronic moral identity has an effect similar to that of accessible moral identity, consistent with prior research (Aquino, McFerran, and Laven 2011; Reed, Aquino, and Levy 2007). Thus, we do not believe that the source of moral identity (chronic or situational) significantly impacts the findings demonstrated here, though a contrasting identity that is situationally accessible may attenuate the effects of chronic moral identity, consistent with Aquino et al. (2009). Taken together, our findings are consistent with existing research such that traits (i.e., chronic moral identity) are malleable because they are multifaceted, whereas temporarily accessible identities have more predictable effects, though for a shorter period of time (Wheeler, DeMarree, and Petty 2010). Managerial Implications Our findings offer several insights for brand managers. First, recognizing that today’s diverse society consists of many outgroups and a small number of ingroups (Pitts and Jarry 2007), improving consumers’ less favorable attitudes toward an out-group brand is an important issue for brand managers. Brands may seek to expand beyond niche submarkets which they originally served to reach a broader consumer base (Aaker 1991; Erdem and Sun 2002) as is the case for brands such as Adidas, Cadillac, and Whole Foods. Our results are likely to be of particular benefit to such brands attempting to reposition to an in-group brand for current out- 7 Chronic moral identity was measured by the internalization dimension of the Self-Importance of Moral Identity Scale (Aquino and Reed 2002).When chronic moral identity was entered in the model with accessible moral identity and brand group membership on brand attitudes, it was not significant nor were there any significant interactions. While some research studies the interaction of chronic and situationally accessible identities (Aquino et al. 2009; Zhang and Khare 2009), this was not the intent of this research. 33 group consumers or those attempting to expand their existing brand image beyond their current in-group consumers. Of course, marketers are unlikely to know the level of any given consumer’s moral identity, making these results difficult to apply. However, we demonstrate that out-group brand attitude may be improved when activating moral identity in an advertisement. While the elicitation of moral identity through an advertisement (Study 3) may be subtle, it is likely to be effective for many consumers given that most consumers hold a base level of morality as important to themself (Aquino and Reed 2002) that just needs to be made accessible. That is, even consumers for whom moral traits are not chronically active, a relatively subtle prime in an advertisement could activate thoughts regarding one’s moral identity, making this identity more relevant than other identities, even if only temporarily. Thus, marketers need to ensure that the communications eliciting moral identity are presented near the point of purchase or at the final stages of the decision-making process (e.g., POP advertising displays or storefront signs eliciting moral identity). Marketers can also appeal to consumers’ chronic moral identity. A benefit of appealing to chronic moral identity to increase out-group brand attitudes is that the proximity of the appeal to the point of purchase may not be as critical such that television commercials or other types of marketing communications (e.g., direct mail, email) may be effective for consumers with chronically high moral identity even when occurring at some point prior to the purchase. Additionally, marketers may think that it is not necessary to elicit moral identity through advertisements for those with chronic moral identity; however, doing so may prevent other identities that are temporarily accessible from overriding the effect of chronic moral identity (Aquino et al. 2009). Future research should consider whether the effectiveness of an 34 advertisement eliciting moral identity depends on consumer characteristics such as chronic moral identity as well as product characteristics such as industry. We also offer insight into how consumers’ morality can interact with brands’ morality (as evidenced through CSR activities; Lichtenstein, Drumwright and Braig 2004; Sen and Bhattacharya 2001), which may be particularly important to brands managers depending on consumers’ current perceptions of their brand’s morality. Though some firms prominently engage in CSR initiatives likely resulting in consumer perceptions of brand morality (e.g., TOMS shoes, Ben & Jerry’s), other brands may be perceived as morally neutral either because they do not engage in CSR activities or simply because consumers are not aware of the brands’ moral initiatives. We demonstrate that accessible moral identity can increase brand attitudes for morally neutral out-group brands such that eliciting moral identity in promotional campaigns may enhance out-group brand attitudes even if brands are not actively engaged in CSR activities or other moral behaviors. In other words, out-group brands need not be perceived as moral to reap the benefits of an accessible moral identity. We also find that consumers have more favorable brand attitudes toward moral brands regardless of accessible identity, which suggests that firms engaging in CSR activities will likely benefit from increasing consumers’ awareness of such activities even if they forgo eliciting moral identity. For example, if a company is not perceived as particularly moral (e.g., offering gambling, cigarettes, weapons, or unhealthy food), they may benefit by increasing consumer awareness of CSR activities in which they are engaged. Examining these effects is an important avenue for future research as the managerial implications may be substantial. To be sure, not all brands desire to improve out-group members’ brand attitudes. Niche alternative brands such as Hot Topic and Irn Bru or luxury brands such as Hermes and 35 Mercedes-Benz may strive to be perceived as out-group brands as this increases the desirability of the brand to the ingroup (Kapferer 2001; Štrach and Everett 2006). Thus, there are some benefits to being perceived as out-group brands. These brands may want to continue to use specific social identities (i.e., student, business professional, soccer mom, etc.) in their advertisements to ensure that the ingroup/outgroup distinction remains. Limitations and Future Research It is important to note that the effect sizes were weaker in studies using real brands compared to those using hypothetical brands, which may be due not only to the variation in selfidentified brands but also to the heterogeneity of associations that consumers have for real brands. To overcome this limitation, Study 3 utilized a hypothetical brand, but, in doing so, failed to obtain the significant main effect of brand group membership. Study 4 also used a hypothetical brand to avoid consumers’ pre-existing perceptions of brand (im)morality, but doing so may have limited the nuances of the joint effect of consumer moral identity and brand morality on out-group brand attitudes if consumers hold strong brand morality perceptions. These are important limitations that advocate for additional research with real brands. Replicating the current findings with real brands is essential to understand whether these effects are strong enough to withstand the realities of the marketplace in which an array of factors influence consumers’ brand attitudes and choices. Future research should also seek to identify additional factors that may attenuate consumers’ unfavorable attitudes toward out-group brands. Though prior research on reference groups has primarily focused on boundary conditions of favorable in-group brand attitudes (see White and Dahl 2007 for an exception), we believe that out-group brand attitudes can be increased in diverse ways, one of which includes appealing to consumers’ moral identity. 36 Additionally, our pilot study revealed that moral identity can influence categorization such that accessible moral identity results in more holistic categorizations. This suggests the possibility that moral identity might come into play for other marketplace judgments. For example, consumers’ perceptions of brand extensions or line extensions might vary depending on their level of moral identity, as this type of assessment involves categorization to a certain extent. Similar to past research which has demonstrated that interdependence, which is associated with more relational, holistic processing, may result in more favorable attitudes toward brand extensions with low fit (Ahluwalia 2008), future research should examine whether the expansiveness associated with moral identity may result in more favorable attitudes toward brand extensions, particularly for those extensions that are otherwise perceived to have low fit. Given that accessible moral identity resulted in flexible categorizations of objects, they may also be more likely to perceive a greater fit between parent and extended brands. In sum, this research opens up the possibility that moral identity influences marketplace judgments such that consumers’ less favorable attitudes toward out-group brands may be improved. Future research should examine when and for whom, either by consumer or brand characteristics, the effectiveness of moral identity accessibility can be maximized, so firms can construct their marketing strategy and advertising campaigns accordingly. We conclude that moral identity may act as a gatekeeper, determining which brands are acceptable to consumers, and marketers may be able to utilize this relationship to their benefit. 37 REFERENCES Aaker, David A. (1991), Managing Brand Equity. New York: The Free Press. _____ and Kevin Lane Keller (1990), “Consumer Evaluations of Brand Extensions,” Journal of Marketing , 54 (January), 27-41. Adweek (2010), “Apple's 'Get a Mac,' the Complete Campaign,” (assessed October 31, 2011) [available at http://www.adweek.com/adfreak/apples-get-mac-complete-campaign-130552]. Ahluwalia, Rohini (2008), “How Far Can a Brand Stretch? Understanding the Role of SelfConstrual,” Journal of Marketing Research, 45 (June), 337-50. Aron, Arthur, Elaine N. Aron, and Danny Smollan (1992), “Inclusion of the Other in the Self Scale and the Structure of Interpersonal Closeness,” Journal of Personality and Social Psychology, 63 (October), 596-612. Aquino, Karl and Americus Reed II (2002), “The Self-Importance of Moral Identity,” Journal of Personality and Social Psychology, 83 (December), 1423-40. _____, Dan Freeman, Americus Reed II, Vivien K. G. Lim, and Will Felps (2009), “Testing a Social-Cognitive Model of Moral Behavior: The Interactive Influence of Situations and Moral Identity Centrality,” Journal of Personality and Social Psychology, 97 (1), 123-141. _____, Brent McFerran, and Marjorie Laven (2011), “Moral Identity and the Experience of Moral Elevation in Response to Acts of Uncommon Goodness,” Journal of Personality and Social Psychology, 100 (4), 730-18. Bandura, Albert (1999), “Moral Disengagement in the Perpetration of Inhumanities,” Personality and Social Psychology Review, 3 (3), 193-209. Bearden, William O. and Michael J. Etzel (1982), “Reference Group Influence on Product and Brand Purchase Decisions,” Journal of Consumer Research, 9 (September), 183-94. Berger, Jonah and Chip Heath (2007), “Where Consumers Diverge from Others: Identity Signaling and Product Domains,” Journal of Consumer Research, 34 (August), 121-34. ______ and Lindsay Rand (2008), “Shifting Signals to Help Health: Using Identity Signaling to Reduce Risky Health Behaviors,” Journal of Consumer Research, 35 (August), 509-18. Blasi, Augusto (1984), “Moral Identity: Its Role in Moral Functioning,” in Morality, Moral Behavior, and Moral Development, ed. William Kurtines and Jacob Gewirtz, New York: Wiley, 128–39. Brewer, Marilynn B. (1991), “The Social Self: On Being the Same and Different at the Same Time,” Personality and Social Psychology Bulletin, 17 (5), 475-82. 38 ______ and Wendi Gardner (1996), “Who Is This ‘We’? Levels of Collective Identity and Self Representations,” Journal of Personality and Social Psychology, 71 (1), 83–93. Brown, Thomas J. and Peter Dacin (1997), “The Company and the Product: Corporate Associations and Consumer Product Responses,” Journal of Marketing, 61 (January), 68-84. Chan, Cindy, Jonah Berger, Leaf Van Boven (2012), “Identifiable but Not Identical: Combining Social Identity and Uniqueness Motives in Choice,” Journal of Consumer Research, forthcoming. Chernev, Alexander, Ryan Hamilton and David Gal (2011), “Competing for Consumer Identity: Limits to Self-Expression and the Perils of Lifestyle Branding,” Journal of Marketing, 75 (3), 66-82. Cohen, Jacob (1992), “A Power Primer,” Psychological Bulletin, 112 (1), 155-9. Daley, Jason (2009), “That’s a Starbucks?” Entrepreneur, November, available at http://www.entrepreneur.com/article/203704 Erdem, Tűlin and Baohong Sun (2002), “An Empirical Investigation of the Spillover Effects of Advertising and Sales Promotions in Umbrella Branding,” Journal of Marketing Research, 39 (November), 408-20. Escalas, Jennifer E. and James R. Bettman (2003), “You Are What They Eat: The Influence of Reference Groups on Consumer Connections to Brands,” Journal of Consumer Psychology, 13 (3), 339-48 ______ and ______ (2005) “Self-Construal, Reference Groups, and Brand Meaning,” Journal of Consumer Research, 32 (December), 378-89. Fournier, Susan (1998), "Consumers and Their Brands: Developing Relationship Theory in Consumer Research," Journal of Consumer Research, 24 (March), 323-73. Gόmez, Ángel, John F. Dovidio, Carmen Huici, Samuel L. Gaertner, and Isabel Cuadrado (2011), “The Other Side of We: When Outgroup Members Express Common Identity,” Personality and Social Psychology Bulletin, 24 (December), 1613-26. Hazan, Cindy and Phillip R. Shaver (1987), “Attachment as an Organizational Framework for Research on Close Relationships,” Psychological Inquiry, 5 (1), 1-22. Hardy, Sam A. (2006), “Identity, Reasoning, and Emotion: An Empirical Comparison of Three Sources of Moral Motivation,” Motivation and Emotion, 30 (3), 207-15. 39 ______, Amit Bhattacharjee, Americus Reed II, and Karl Aquino (2010), “Moral identity and Psychological Distance: The Case of Adolescent Parental Socialization,” Journal of Adolescence, 33, 111-23. Herzenstein, Michal, Scott Sonenshein, and Utpal M. Dholakia (2011), “Tell Me a Good Story and I May Lend You My Money: The Role of Narratives in Peer-to-Peer Lending Decisions”, Journal of Marketing Research, 48 (Special Issue), S138-49. Isen, Alice M. and Kimberly A. Daubman (1984), “The Influence of Affect on Categorization,” Journal of Personality and Social Psychology, 47 (December), 1206-17. Jones, Bryan and Howard Rachlin (2006), “Social discounting,” Psychological Science, 17 (4), 283-86. Kapferer, Jean-Noel (2001), Reinventing the Brand: Can Top Brands Survive the New Market Realities?, Kogan Page, London. LeBoeuf, Robyn A., Eldar Shafir, and Julia Belyavsky Bayuk (2010), “The Conflicting Choices of Alternating Selves,” Organizational Behavior and Human Decision Processe, 111 (1), 48- 61. Liberman, Varda, Steven M. Samuels, and Lee Ross (2004), “The Name of the Game: Predictive Power of Reputations versus Situational Labels in Determining Prisoner’s Dilemma Game Moves,” Personality and Social Psychology Bulletin, 30 (9), 1175-85. Lichtenstein, Donald R., Minette Drumwright, and Bridgette Braig (2004), “The Effect of Corporate Social Responsibility on Customer Donations to Corporate‐Supported Nonprofits,” Journal of Marketing, 68 (October), 16-32. McFerran, Brent, Darren W. Dahl, Gavan J. Fitzsimons, and Andrea C. Morales (2010), “I’ll Have What She’s Having: Effects of Social Influence and Body Type on the Food Choices of Others,” Journal of Consumer Research, 36 (6), 915-29. Mikulincer, Mario, Peri Kedem, Dov Paz (1990), “Anxiety and categorization - 1. The structure and boundaries of mental categories,” Personality and Individual Differences, 11 (8), 805-14. Pitts, Davis and Elizabeth M. Jarry (2007), “ Ethnic Diversity and Organizational Performance: Assessing Diversity Effects at the Managerial and Street Levels,” International Public Management Journal, 10(2), 233‐54. Preacher, Kristopher J., Derek D. Rucker, and Andrew Hayes (2007), “Addressing Moderated Mediation Hypotheses: Theory, Methods, and Prescriptions,” Multivariate Behavioral Research, 42, 185-227. 40 Reed, Americus II and Karl Aquino (2003), “Moral Identity and the Expanding Circle of Moral Regard toward Out-Groups,” Journal of Personality and Social Psychology, 84 (June), 1270- 86. ______, ______, and Eric Levy (2007), “Moral Identity and Judgments of Charitable Behaviors,” Journal of Marketing, 71 (January), 178-93. Rosch, Eleanor H. (1975), “Cognitive Representations of Semantic Categories,” Journal of Experimental Psychology: General, 4 (3), 192-233. Rosenthal, Harriet E. S., and Richard J. Crisp (2006), “Reducing stereotype threat by blurring intergroup boundaries,” Personality and Social Psychology Bulletin, 32, 501-11. Sen, Sankar and C.B. Bhattacharya (2001), “Does Doing Good Always Lead to Doing Better? Consumer Reactions to Corporate Social Responsibility,” Journal of Marketing Research, 38 (May), 225-43. Skarlicki, Daniel P, Danielle D. van Jaarsveld, and David D. Walker (2008), Getting Even for Customer Mistreatment: The Role of Moral Identity in the Relationship Between Customer Interpersonal Injustice and Employee Sabotage,” Journal of Applied Psychology, 93 (6), 1335-47. Sood, Sanjay, and Kevin Lane Keller (2012), “The Effects of Brand Name Structure on Brand Extension Evaluations and Parent Brand Dilution,” Journal of Marketing Research, 49 (June), 373-82. Štrach, Pavel and André M. Everett (2006), “Brand Corrosion: Massmarketing’s Threat to Luxury Automobile Brands after Merger and Acquisition”, Journal of Product and Brand Management, 15 (2), 106-20. Swaminathan, Vanitha, Karen Stilley, and Rohini Ahluwalia (2009), “When Brand Personality Matters: The Moderating Role of Attachment Styles,” Journal of Consumer Research, 35 (April), 985-1002. Tajfel, Henri and John C. Turner (1979), “An Integrative Theory of Intergroup Conflict,” The Social Psychology of Intergroup Relations, Monterey, CA: Brooks/Cole. Trope, Yaacov and Nira Liberman (2003), “Temporal Construal,” Psychological Review, 110 (3), 403-21. ______ and ______ (2010), “Construal Level Theory and Psychological Distance,” Psychological Review, 117 (2), 440-463. 41 Turner, John C. (1985), “Social Categorization and The Self- Concept: A Social Cognitive Theory of Group Behavior”, in E. J. Lawler (ed.), Advances in Group Processes: Theory and Research, 2, 77-121. Ülkümen, Gülden, Amitav Chakravarti, and Vicki G. Morwitz (2010), “Categories Create MindSets: The Effect of Exposure to Broad versus Narrow Categorizations on Subsequent, Unrelated Decisions,” Journal of Marketing Research, 47 (August), 659-71. Verrochi, Nicole, Americus Reed II, and Jennifer Tong (2009), “Moral Identity and Attributions of Corporate Social Responsibility,” Advances in Consumer Research, 36, 51-52. Wheeler, S. Christian, and Jonah Berger (2007), “When the Same Prime Leads to Different Effects,” Journal of Consumer Research, 34 (October), 357-68. ______, Kenneth G. DeMarree, and Richard E. Petty (2008), “A Match Made in the Laboratory: Persuasion and Matches to Primed Traits and Stereotypes,” Journal of Experimental Social Psychology, 44, 1035-47. White, Kate and Darren Dahl (2006), “To Be or Not Be: The Influence of Dissociative Reference Groups on Consumer Preferences,” Journal of Consumer Psychology, 16 (4), 404-13. ______ and ______ (2007), “Are All Out‐Groups Created Equal? Consumer Identity and Dissociative Influence,” Journal of Consumer Research, 34 (December), 525-36. ______ and Jennifer J. Argo (2011), “When Imitation Doesn’t Flatter: The Role of Consumer Distinctiveness in Responses to Mimicry,” Journal of Consumer Research, 38 (December), 667-80. Winterich, Karen P., Vikas Mittal, and William T. Ross Jr. (2009), “Donation Behavior toward In‐Groups and Out‐Groups: The Role of Gender and Moral Identity,” Journal of Consumer Research, 36 (August), 199-214. Yoon, Carolyn, Angela H. Gutchess, Fred Feinberg, and Thad A. Polk (2006), “A Functional Magnetic Resonance Imaging Study of Neural Dissociations between Brand and Person Judgments,” Journal of Consumer Research, 33 (June), 31-40. Zhang, Yinlong and Adwait Khare (2009), “The Impact of Accessible Identities on the Evaluation of Global versus Local Products,” Journal of Consumer Research, 36 (October), 524-37. 42 Table 1. Means for Inclusion of Other in the Self (IOS) by Brand Group Membership and Accessible Identity (Studies 3 and 4) Accessible Identity Brand Group Membership In-group brand Out-group brand Study 3 BI condition 2.23 2.00 MI condition 2.14 2.81 Study 4 (Neutral condition only) SI condition 3.07 1.88 MI condition 3.43 3.73 43 Figure 1. Effect of Accessible Identity and Brand Membership on Brand Attitude (Study 1) 4.03 6.61 5.14 6.40 1 2 3 4 5 6 7 Out-group brands In-group brands SI condition MI condition Brand Attitude 44 Figure 2. Effect of Accessible Identity and Brand Membership on Brand Attitude (Study 2) 2.38 6.28 6.17 3.14 6.10 6.04 1 2 3 4 5 6 7 Dissociative-group brand In-group brand Aspirational-group brand SI condition MI condition Brand Attitude 45 Figure 3. Effect of Accessible Identity and Brand Membership on Brand Attitude (Study 3) 4.51 5.18 5.25 4.80 1 2 3 4 5 6 7 Out-Group Brand In-Group Brand BI condition MI condition Brand Attitude 46 Figure 4. Effect of Brand Morality, Accessible Identity, and Brand Membership on Brand Attitude (Study 4) A. Moral Brand B. Neutral Brand C. Immoral Brand 4.56 5.13 4.70 4.86 1 2 3 4 5 6 7 Out-Group Brand In-Group Brand SI condition MI condition Brand Attitude 3.96 4.71 4.97 4.31 1 2 3 4 5 6 7 Out-Group Brand In-Group Brand SI condition MI condition Brand Attitude 3.57 3.71 3.02 3.57 1 2 3 4 5 6 7 Out-Group Brand In-Group Brand SI condition MI condition Brand Attitude 47 Appendix A. Description of Study Measures 1. Identity Prime Manipulation Check (Study 1, 2, 4, from Reed, Aquino, and Levy 2007) Please indicate the extent to which your writing reflects how you see yourself as: 1) a student, 2) a member of an organization, 3) a moral person, and 4) safety conscious (to some extent = 1, to a great extent = 7). In the second pretest for Study 3, safety conscious was replaced with a business leader and “your writing” was replaced with “this advertisement”. 2. Brand Attitude (Study 1-4, from Sood and Keller 2012; White and Dahl 2007) An attitude toward each brand was assessed with four items on a 7-point scale anchored by: bad/good, unfavorable/favorable, negative/positive, and low quality/high quality. 3. Advertisement Evaluation (Study 3 Pretests 1 and 2, Main Study 3) Please indicate your evaluation of the advertisement of Mardi: 1) very unfavorable/very favorable and 2) dislike very much/like very much (on a 7-point scale). 4. Quality/Brand Morality (Study 3 Pretest 2) Please indicate the quality and morality of the Mardi brand on two items (very low quality/very immoral = 1, very high quality/very moral = 7). 5. Psychological Distance (Study 3, 4, from Aron, Aron, and Smollan 1992) Please indicate which set of overlapping circles below best represents how you feel regarding (brand name). 6. Brand Morality (Study 4, from Aquino and Reed 2002 Moral Identity Instrument) Please indicate to what extent the following characteristics describe a watch brand, “Mardi” (caring, fair, friendly, generous, and kind; 1 = not at all; 7 = very much). 7. Newspaper Article Check (Study 4 Pretest) How realistic/relevant was the newspaper article? (1 = not at all; 7 = very much). 48 Appendix B. Advertisements Priming Moral Identity and Business Identity A: Moral Identity Condition B: Business Identity Condition Being caring, compassionate, fair, friendly, generous, helpful, hardworking, honest an d kind… As much as you value these sprits, do we O ld Navy either. With these sprits, we proudly launch our n You are a person who is caring, compassionate, fair, friendly, generous, helpful, hardworking, honest and kind. For people such as you – Mardi watches. Being caring, compassionate, fair, friendly, generous, helpful, hardworking, honest an d kind… As much as you value these sprits, do we O ld Navy either. With these sprits, we proudly launch our n You are a business professional, who has high standards, exceeds expectations, and is committed to success. For business professionals like you –Mardi watches. 49 Appendix C. Newspaper Articles Presented in Study 4 All participants read the following paragraph, which ended with one of the three randomlyassigned brand morality conditions. For the fourth year in a row, several states like Maryland and California have topped the nation in the proportion of high school graduates who successfully passed the rigorous Advanced Placement exams, suggesting that the commitment of these states to investment in education is paying off in with the kind of advanced, high-level academic skills necessary to compete successfully. However, the huge gap between the highest and lowest performing schools tracks almost exactly the differences in wealth between the state's richest and poorest jurisdictions, which is great cause for concern. A. Moral Brand Condition Mardi, the watch brand, has recently launched a special initiative aimed at thwarting the widening educational gap between the rich and poor. Mardi says they will donate $2 for each Mardi watch sold in the US. Representatives of Mardi say educating America’s future is a top priority. B. Immoral Brand Condition Recently, companies from several large consumer brands, including Mardi, held a meeting to discuss ways to address the widening educational gap between the rich and poor. Several companies announced that they will launch a special initiative to improve education by donating $2 for each of their products sold in the US. However, reports state that Mardi, the watch brand, has decided not to participate in this initiative since it may decrease profits. Representatives from Mardi could not be reached for comment. C. Neutral Brand Condition Mardi, the watch brand, is considering launching a special initiative aimed at thwarting the widening educational gap between the rich and poor. At this time, representatives from Mardi did not specify if or when such a program would begin. Note: These manipulations were developed from an actual newspaper article that was adapted to influence brand morality. 50 Web Appendix PILOT STUDY Procedure Seventy-five undergraduates (65% male) participated in the study in exchange for course credit. First, participants completed the essay writing task used in Studies 1, 2, and 4 for moral identity (MI) or student identity (SI) and responded to the manipulation check items. Then, in an ostensibly unrelated study, participants categorized eight clothing items and eight vegetables used in Rosch (1975) and Isen and Daubman (1984). For each category, there were two excellent exemplars (shirts and pants for clothing; peas and carrots for vegetable), two medium exemplars (stockings and sandals for clothing; tomatoes and artichoke for vegetable), and four weak exemplars (purse, rings, cane and gloves for clothing; mushrooms, potatoes, green onions and pumpkin for vegetable). The first item for the category was always an excellent exemplar with the seven remaining items presented in random order. Participants rated the extent to which each item belonged to the given category on a 10- point scale (1 = definitely does not belong to the category, 5 = does not belong to the category but is very similar to members of that category, 6 = does belong to the category but is not a very good example of it, 10 = definitely belongs to the category). We expected differences for the weak and medium exemplars, but not the excellent exemplars (Isen and Daubman 1984). Since there were no differences in the effect of identity by category type (i.e., vegetables or clothing), we collapsed the prototypicality ratings into the eight weak exemplars (α = .65; four per category), the four medium exemplars (α = .22; two per category) and the four strong exemplars (α = .67; two per category). We note that reliability for the four medium exemplars is low, but we believe this is due to the inherent variation in the categorization of medium exemplars. 51 Results Manipulation Check. As expected, participants in the MI condition indicated their writing reflected them as a moral person more than those in the SI condition (MMI = 5.42 vs. MSI = 4.51; F(1,73) = 7.20, p < .01). In contrast, those in the SI condition indicated their writing reflected them as a student more than those in the MI condition (MMI = 4.87 vs. MSI = 5.97; F(1,73) = 15.21, p < .001). Thus, the identity primes were successful. Categorization. To determine whether MI would result in broader categorizations such that weaker exemplars would be perceived as more similar to a given category, we conducted a repeated-measures ANOVA with exemplar type (weak, moderate, excellent) as a within-subjects factor and identity condition as a between-subjects factor. As expected, there was a main effect of exemplar type (F(2,146) = 259.90, p < .001; Mhigh = 9.59. Mmedium = 7.32, Mweak = 6.11). The main effect of identity condition was also significant (F(1,73) = 14.99, p < .0001). Moreover, the results revealed a significant interaction between exemplar type and identity condition (F(2,146) = 7.13, p < .01). The effect size of the interaction of exemplar type and identity condition was large (Cohen 1992; =.089; Cohen’s d = .625). Further analysis revealed that the effect of identity was not significant for excellent exemplars (MSI = 9.43, MMI = 9.75; t(73) = -1.35, p > .18), as expected and consistent with Isen and Daubman (1984). However, the effect of identity was significant for medium exemplars (MSI = 6.95, MMI =7.67; t(73) = -2.03, p = .04) as well as for weak exemplars (MSI = 5.37, MMI = 6.84; t(73) = -6.12, p < .01) such that those with accessible MI perceived weak and moderate exemplars to belong to a given category to a greater extent than those with accessible SI. These results are consistent with our prediction that an accessible MI results in broader, more inclusive categorization of objects even when the objects are unrelated to morality.

can brands move in from the outside.pdf

0.57MB

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

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