Psych DB

677

The Journal of Social Psychology, 2009, 149(6), 677–693 Copyright © 2009 Heldref Publications

VSOC0022-4545The Journal of Social Psychology, Vol. 149, No. 6, Oct 2009: pp. 0–0The Journal of Social Psychology

Culture and Individuation: The Role of Norms and Self-Construals

The Journal of Social Psychology Boucher & Maslach

HELEN C. BOUCHER Bates College

CHRISTINA MASLACH University of California, Berkeley

ABSTRACT. Despite mounting evidence that members of Asian cultures are less likely to engage in behavior that makes them appear distinctive (i.e., individuating behavior) than members of prototypical Western culture, the direct mechanisms through which this effect occurs have not been explored. In the present research, we examined the role of judgments of social appropriateness for both acceptance of and comfort engaging in individuating behavior, and the role of the divergent self-construals thought to characterize these cultural groups (i.e., Asian and Euro-Americans). Results indicated that ethnicity moder- ated the relationship between judgments of social appropriateness and acceptance of indi- viduating behavior, and self-construals mediated the relationship between ethnicity and both acceptance of and comfort engaging in individuating behavior. Limitations and implications are discussed.

Keywords: cultural differences, ethnicity, individuation, norms, self-construals

THE NEED FOR PERSONAL UNIQUENESS or distinctiveness has been postulated to be both an important social value and even a fundamental human need (Vignoles, Chryssochoou, & Breakwell, 2000). Although this drive is tem- pered by a competing drive for similarity, individuals express a need for unique- ness on self-report measures, overestimate the uniqueness of their positive characteristics, and experience discomfort and act in ways to re-establish a moderate degree of distinctiveness when made to feel too similar to others (Lynn & Snyder, 2002). However, all behavior is judged according to the context in which it occurs. An important part of this context, and one now receiving the

The research reported herein was submitted in partial fulfillment of the first author’s master’s thesis. The authors wish to thank Kaiping Peng and Ann Kring for their helpful comments on earlier versions of this manuscript.

Address correspondence to Helen C. Boucher, Bates College, Department of Psychology, Lewiston, ME 04240, USA; [email protected] (e-mail).

678 The Journal of Social Psychology

attention it is due in social psychology, is the larger cultural context in which behavior takes place. The same action may have different meanings and func- tions depending on the culture in which it is performed (Markus & Kitayama, 1991). This may be especially true of behavior that makes one appear unique or distinctive from others (i.e., individuating behavior). Cultures likely differ in the extent to which they encourage individuating behavior, with the result that the same individuating behavior is accepted and even encouraged in one culture but criticized and condemned in another. The purpose of this research is to examine the relationship between culture and individuation. Specifically, we tested whether Euro-Americans and Asian Americans differ in both their willingness to engage in individuating behavior and their reactions to the individuating actions of others. We also examined how judgments of the social appropriateness (i.e., normativeness) of individuating behavior factor into these responses. Finally, we examined whether the divergent self-construals of these groups mediate these differences, if they exist.

Culture and Individuation: A Brief Review

Individuation can be conceived as both a situationally induced state or as an individual difference; that is, one can feel differentiated from others (whether one wants to be or not) due to contextual factors, or be chronically willing to stand out. Maslach (1974) provided an early demonstration of the former conceptual- ization, reporting that individuals were more likely to individuate themselves when they expected positive rather than negative outcomes. Somewhat later, Maslach and her colleagues (1985) developed an individual difference measure of a person’s willingness to individuate her- or himself. As expected, people scoring highly on the Individuation Scale reported a greater frequency of owning distinctive possessions, expressing unique opinions, and so on (Maslach, Stapp, & Santee, 1985). High individuators were more likely to disagree with the opin- ions of other group members, especially when the others agreed on their opinion (Maslach, Santee, & Wade, 1987). They also have more impact in social situa- tions, as seen in greater creativity and leadership (Whitney, Sagrestano, & Maslach, 1994). From this, one gets the highly positive picture of individuators as natural leaders: They are assertive yet open-minded, creative, expressive indi- viduals who are not afraid to stand out and buck the system.

Yet, it is important to keep in mind that these studies were conducted in the United States, an individualistic culture that prizes assertive self-expression and the pursuit of individuality. Individuation in the U.S. is likely part of a cultural imperative to find and assert one’s unique identity. However, collectivistic cultures stress the importance of group harmony and fitting in with others; thus, it is reasonable to expect that individuation is less likely to occur in these cultures and, when it does occur, is likely to be socially sanctioned (Markus & Kitayama, 1991). In support of this, some research indicates that members of collectivistic

Boucher & Maslach 679

cultures are more conforming than members of individualistic ones (Bond & Smith, 1996), and relatively allocentric (i.e., collectivistic, measured at the indi- vidual level) individuals scored lower on an individual difference measure of need for uniqueness than relatively idiocentric individuals (Yamaguchi, Kuhlman, & Sugimori, 1995). Kim and Markus (1999) argue that uniqueness has positive connotations of freedom and independence while conformity has conno- tations of societal oppression and passivity in the U.S.; at the same time, unique- ness and conformity have negative connotations of deviance and positive connotations of connectedness and harmony, respectively, in East Asia. Thus, Americans are taught to transcend and even break norms and East Asians to follow them. Indeed, they found that Euro-Americans had more unique prefer- ences than both Chinese Americans and Koreans (e.g., they chose a relatively uniquely-colored pen as a gift for completing a questionnaire). Also, American ads are more likely to emphasize themes of uniqueness, while Korean ads emphasize themes of conformity (Kim & Markus, 1999).

More direct is evidence using the Individuation Scale, and similar conclu- sions can be drawn: Asian Americans score lower than Euro-Americans (Maslach et al., 1985). Kwan and her colleagues (2000) found that the Individua- tion Scale has a different factor structure in Hong Kong than the 1-factor struc- ture reported by Maslach et al. (1985). In this relatively collectivistic culture, the scale broke down into two factors, one that captures Taking the Lead behaviors of a Chinese leader (e.g., volunteering to head a committee, performing on a stage before a large audience), and one that captures inappropriate behaviors undertaken simply to Seek Attention and not for any constructive social purpose (e.g., challenging a speaker whose opinion clashes with one’s own, speaking up about one’s ideas despite being uncertain whether they are correct). As expected, Taking the Lead items were rated as more desirable than Seeking Attention items. Thus, these two types of behavior have the same meaning in the U.S. (hence the 1-factor model) but take on different meanings in a collectivistic culture and are better represented with two factors (Kwan, Bond, Boucher, Maslach, & Gan, 2000).

Similarly, Tafarodi, Marshall, & Katsura (2004) argue that an important factor to consider when making predictions about the relationship between cul- ture and distinctiveness is the type of distinctiveness being discussed. Specifi- cally, people living in collectivistic cultures will be less likely to want to stand out through unique choices in possessions, hobbies, and the like, but will be willing to stand out in ways that connote the achievement of merit or excellence (e.g., attaining a promotion at work more quickly than is the norm). These kinds of distinctiveness seem to map squarely onto Kwan et al.’s (2000) Seeking Attention and Taking the Lead, respectively. In their research, they found that Japanese expressed less desire to appear different for its own sake (i.e., when it was not tied to achievement or merit), and experienced this type of distinctive- ness as less positive than Euro-Canadians.

680 The Journal of Social Psychology

While interesting, missing from past research are direct tests of the mecha- nisms responsible for documented ethnic/national differences. For example, Kim & Markus (1999) argue that cultural differences in self-construals explain ethnic differences in preferences for uniqueness and conformity. That is, since people in the U.S. have an independent self-construal, they prefer uniqueness, and since people in Korea have an interdependent self-construal, they prefer conformity. However, they provided no direct evidence for this hypothesized mediator. Simi- larly, Kim & Markus (1999), Kwan et al. (2000) and Tafarodi et al. (2004) argue that a heightened awareness of norms underlies attitudes and emotional reactions concerning the pursuit of distinctiveness. But then again, they provided no direct evidence. Cross-cultural psychologists have been criticized for failing to provide direct evidence for the hypothesized mediators and moderators of their effects (Matsumoto, 1999). To redress this, the present research focused on the role of both self-construals and judgments of the social appropriateness of individuating behavior.

Self-Construal: Mediating the Link Between Ethnicity and Individuation?

Self-construals, or characteristic ways of viewing the self, is one of the most prominent constructs invoked to explain cultural variation (Markus & Kitayama, 1991). People from more individualistic cultures (including the U.S.) tend to have an independent self-construal; that is, the self is separate from social context, bounded, and stable. One’s internal attributes are self-defining (e.g., traits, feelings), and goals include striving to be unique, expressing oneself, and promoting one’s own goals. In contrast, people from more collectivistic cultures (including Eastern and Southern Asia) tend to have an interdependent self-construal; that is, the self is connected to others and to social context, flexible, and variable. One’s external attributes are self-defining (e.g., statuses, roles), and salient goals include fitting in, engaging in appropriate action, and promoting the goals of others. Others take on central importance, as it is relationships with others in specific contexts that define the self (Markus & Kitayama, 1991).

While Markus and Kitayama have been criticized for reporting ethnic/ national differences and attributing them to self-construals without including an actual measurement of them (Matsumoto, 1999), other researchers have devel- oped measures of the construct. For example, Singelis (1994) developed a Self- Construal Scale to measure the strengths of a person’s independent and interdepen- dent self-construals. However it is measured, the implication of self-construals for individuation is clear: Since part of the cultural imperative for people with an independent self-construal is to be unique, and since this is exactly what happens when one is individuated relative to others, people with an independent self-construal should be more likely to express a willingness to individuate themselves than people with an interdependent self-construal, especially in the case of actions that amount to getting attention for its own sake. In addition, those with an independent

Boucher & Maslach 681

self-construal should be more accepting of the Seeking Attention behavior of others than those with an interdependent self-construal. Furthermore, given the analysis above, people with either an independent or interdependent self- construal should be willing to engage in and accepting of individuating actions that connote leadership or achievement (i.e., Taking the Lead behavior).

It is necessary to mention an important conceptual refinement of self-construals at this point. Although Singelis’ (1994) Self-Construal Scale was published as two factors (independent and interdependent), others have maintained that the scale is best represented as several independent and interdependent facets. Specifically, Hardin, Leong, & Bhagwat (2004) found that the independence factor broke down into four facets, labeled autonomy/assertiveness (“auton- omy”), individualism, behavioral consistency, and primacy of self, and interde- pendence broke down into group esteem and relational interdependence. In subsequent work, Hardin (2006) found that specific facets were better predictors of social anxiety than the subscales (e.g., for Asian Americans, greater social anxiety was associated with lower autonomy, lower behavioral consistency, and higher relational interdependence). This research runs parallel to an influential meta-analysis of measures of individualism/collectivism (Oyserman, Coon, & Kemmelmeier, 2002), in which individualism broke down into six and collectiv- ism into eight domains. Importantly, the authors determined that the presence or absence of cultural differences in individualism/collectivism often hinges on which domain(s) are used.

We thought that given this, specific facets would be more highly related to individuation than the subscales and, after examining the items from each facet, determined it most likely that the autonomy facet of independence would be positively related and the group esteem facet of interdependence negatively related to comfort engaging in and acceptance of individuating behavior. That is, autonomy items (e.g., “I prefer to be direct and forthright when dealing with people I’ve just met”) refer to charting one’s own course in life without feeling beholden to others’ expectations, which seems to bear the closest relationship conceptually with individuating behavior. Individualism items (e.g., “My personal identity, independent of others, is very important to me”) reflect self- definition as a bounded individual, primacy of self items (e.g., “Being able to take care of myself is a primary concern for me”) reflect putting one’s own goals over others’ goals, and behavioral consistency items (e.g., “I act the same way no matter who I am with”) refer to having a consistent self-concept across contexts, and while these three facets are clearly part of the independence construct (Markus & Kitayama, 1991), they do not seem as closely related to individuation. By the same token, group esteem items (e.g., “It is important for me to maintain harmony within my group”) refer to respecting authority and striving to maintain group harmony, while relational interdependence items (e.g., “If my brother or sister fails, I feel responsible”) reflect putting close others’ goals in front of one’s own. Theoretically, non-individuation is desirable in part because individuation

682 The Journal of Social Psychology

can be disruptive; thus, we thought that group esteem would be more closely related to a desire to not individuate oneself than relational interdependence. However, given both the popularity of using the whole subscales to test for medi- ation, and previous demonstrations of mediation by the subscales (see Lam & Zane, 2004, for an example of the mediating effect of independence on ethnic differences in coping via primary control), we first examined whether indepen- dence and interdependence mediated ethnic differences in individuation, and then tested the impact of the autonomy and group esteem facets.

The Role of Social Appropriateness

We also examined the role that judgments of social appropriateness (or nor- mativeness) play in both willingness to individuate oneself and reactions to oth- ers’ individuating behavior. Kim and Markus (1999) suggested, for example, that in the U.S. “people follow the norm not to follow norms” (p. 787), while in East Asia “people follow the norm to follow norms” (p. 786; Ybarra & Trafimow, 1998). Building on this idea, it is likely that the relationship between judgments of normativeness and willingness to engage in, and acceptance of, individuating behavior is stronger for Asian Americans than Euro-Americans. That is, Asian Americans should be more likely to report comfort engaging in, and acceptance of, individuating behavior to the extent that it is socially appropriate.

In the present research, we compared individuals of European and Asian descent, all living in the U.S. We thought this would be an especially interesting comparison to make, given growing evidence that Asian Americans are bicultural; that is, have been heavily exposed to both individualistic and collectivistic values (Hong, Morris, Chiu, & Benet-Martínez, 2000). Thus, using this group provides an even stronger test of our hypotheses, to the extent that it has been exposed to indi- vidualistic models regarding individuation. Additionally, using Asian Americans avoids a problem that plagues cross-cultural research. For example, sometimes cross-cultural researchers report that Japan is more individualistic than the U.S., when theoretically the opposite should be true (Matsumoto, 1999). This may in part be explained by the reference-group effect, or the idea that participants make comparisons with relevant others when asked to rate themselves on things such as values (Heine, Lehman, Peng, & Greenholtz, 2002). Thus, a Japanese person may rate her- or himself as quite individualistic compared to the other Japanese people around her or him. Using Asian Americans (especially at a diverse campus such as U.C. Berkeley, where the present research was conducted) eliminates, or at least greatly reduces, this effect (Heine et al., 2002).

Summary of Hypotheses

Replicating previous research, we expected that Euro-Americans would score more highly on the Individuation Scale than Asian Americans. In addition,

Boucher & Maslach 683

we expected that ethnic differences would be more apparent in the individuating behavior labeled Seeking Attention by Kwan et al. (2000). We made no a priori prediction regarding Taking the Lead behavior, but anticipated that either there would be no differences, or Euro-Americans would be more comfortable engag- ing in this behavior as well. Importantly, we predicted that Asian Americans would feel more comfortable engaging in Taking the Lead actions than Seeking Attention ones. To the extent that Taking the Lead behavior is seen as more socially appropriate, this would provide evidence that Asian Americans more carefully consider the normativeness of individuating actions when thinking about engaging in them or not.

To more directly examine the effect of social appropriateness, we developed several scenarios depicting individuating behavior and had our participants rate them for social appropriateness, comfort in performing them, and personal acceptance of them. We anticipated that Euro-Americans would express more comfort engaging in and acceptance of these behaviors. Importantly, we predicted that the relationship between judgments of social appropriateness and willingness to engage in, and acceptance of, the individuating behaviors would be stronger for Asian Americans than for Euro-Americans (i.e., the relationship between judgments of social appropri- ateness and scenario responses would be moderated by ethnicity). Finally, we pre- dicted that ethnic differences on both subscales of the Individuation Scale, and comfort and acceptance of the scenarios, would be mediated by the autonomy and group esteem facets of independent and interdependent self-construals.

METHOD

Participants and Procedure

Sixty-nine self-identified Euro-American (35 female, Mage = 21.6) and 76 Asian American (50 female, Mage = 20.1) undergraduates at the University of California, Berkeley, participated in partial fulfillment of psychology course requirements. Of the Asian Americans, there were 47 Chinese, 10 Koreans, 9 Pacific Islanders, 6 Southeast Asians, 2 Japanese, 1 Malaysian, and 1 Taiwanese participant. The majority of the Asian American sample was “second genera- tion,” in that 91% reported that their parents were born somewhere other than the U.S. (compared to 22% of the Euro-Americans). Seventy-five percent of the Asian Americans reported that English was their primary language, compared to 97% of the Euro-Americans. An ANCOVA examining ethnic differences in length of living in the U.S. (controlling for age) revealed that the Euro-Americans (M = 20.2) had lived in the U.S. slightly longer than the Asian Americans (M = 18.0), F(1,142) = 8.83, p < .01, hp

2 = .06.1

Participants completed all measures with four to 14 others. After agreeing to participate, they responded to the scenarios, and then completed the Individua- tion Scale, the Self-Construal Scale, and the demographics measure. No one

684 The Journal of Social Psychology

needed more than an hour to complete everything. After finishing, participants were debriefed and thanked.

Materials

Individuation Scale (12 items, a = .85; Maslach et al., 1985). Participants rated their willingness to engage in various individuating actions on 1 = not at all to 5 = very much scale. Higher means indicate a greater willingness to individu- ate oneself. We also created Taking the Lead (6 items, a = .76) and Seeking Attention (6 items, a = .75) subscales based on the factors identified by Kwan et al. (2000; see p.3 for examples).

Scenarios. We created six scenarios depicting individuating actions varying in social appropriateness (e.g., revealing oneself to be gay during a class discus- sion on prejudice, challenging a professor during a class lecture). A pilot sample of participants (n = 44, comprised of 23 Asian Americans, 15 Euro-Americans, 2 Latinos, 1 Mexican, 2 reporting as multiethnic, and 1 who did not report their ethnicity) read each scenario and rated how much the action would make the person performing it stand out from others, how typical of most people the action was, and how appropriate most people would say the action is on a scale 1 = not at all to 5 = extremely. There was no effect of ethnicity on these ratings (ps = .12, .21, and .20, respectively), and participants agreed that the scenario actions would make the actor stand out (M = 4.28) and were less typical of most people (M = 2.53). One-sample t-tests revealed that both of these means differed signifi- cantly from the scale midpoint, t(43) = 19.21, p < .001, d = 2.91 and t(43) = −5.30, p < .001, d = 0.80, respectively. Thus, we can be confident that the actions depicted in the scenarios were perceived as individuating by the study partici- pants. Also, the scenarios varied substantially in how appropriate the action was considered to be, ranging from M = 1.68 to 4.35. In the present study, participants read each scenario and indicated a) how appropriate most people would say the action is, b) how comfortable the participant her- or himself would be engaging in the action, and c) how accepting the participant her- or himself would be of someone else performing the action, on a scale 1 = not at all to 7 = extremely.

Self-Construal Scale (24 items, independent a = .74, interdependent a = .70; Singelis, 1994). This scale measures the strengths of an individual’s independent and interdependent self-construals with 12 items each. Participants rated their agreement with each item on a 1 = strongly disagree and 7 = strongly agree scale. Higher means indicate a stronger self-construal. We also computed sub- scales from the facets identified by Hardin et al. (2004); specifically, the indepen- dence facet of autonomy (6 items, a = .60), and the interdependence facet of group esteem (6 items, a = .58, see sample items on p.5).2

Demographics. Participants reported their age, sex, ethnic background, pri- mary language, whether their parents were born in the U.S. and how long they had lived in the U.S.

Boucher & Maslach 685

RESULTS

We entered sex as a factor in all analyses, and note effects of sex when they occurred. See Table 1 for means and standard deviations by culture, F values, df, p values, and partial eta squared (hp

2) for all measures. All analyses are two-tailed. First, we conducted analyses on the Individuation Scale. As expected, Euro- Americans scored more highly on the full scale than Asian Americans (see Table 1). Ethnic differences on the two individuation subscales were analyzed in a 2 (ethnicity: Euro-, Asian American) × 2 (subscale type: taking the lead, seeking attention) mixed-model ANOVA, with the former variable as a between-subjects factor and the latter as a within-subjects one. This revealed a main effect of ethnicity (see above), a main effect of subscale type, F(1,143) = 11.48, p < .01, hp

2 = .07, such that Taking the Lead actions (M = 3.19) were rated more highly than Seeking Attention ones (M = 3.01), and a marginal ethnicity by subscale type interaction, F(1,143) = 3.24, p = .07, hp

2 = .02. Given our a priori hypothe- ses, we explored this interaction and found that Euro-Americans scored more highly than Asian Americans on both the Taking the Lead and Seeking Attention

TABLE 1. Summary of Measures for Euro- and Asian Americans

Euro Americans

Asian Americans

Measure M M F df p

Individuation Scale 3.36 (0.68) 2.84 (0.66) 21.50 144 .001 0.13 Taking the Lead 3.40 (0.73) 2.98 (0.83) 10.31 144 .01 0.07 Seeking Attention 3.31 (0.75) 2.70 (0.63) 28.12 144 .001 0.16

Self-Construal Scale Independent subscale 4.76 (0.88) 4.50 (0.66) 4.11 144 .05 0.03 Autonomy 4.57 (1.07) 3.96 (0.70) 16.86 144 .001 0.11 Individualism 5.73 (0.82) 5.65 (0.84) .31 144 .58 0.00 Behavioral consistency

3.66 (1.62) 3.40 (1.44) 1.03 144 .31 0.01

Primacy of self 5.01 (1.09) 5.25 (1.18) 1.55 144 .22 0.01 Interdependent subscale

4.56 (0.68) 5.00 (0.67) 15.44 144 .001 0.10

Group esteem 4.89 (0.77) 5.26 (0.76) 8.55 144 .01 0.06 Relational interdependence

4.49(1.24) 4.80 (0.99) 2.75 144 .10 0.02

Individuating Scenarios Comfort 2.93 (0.94) 2.54 (0.93) 6.32 144 .05 0.04 Acceptance 4.69 (0.69) 4.33 (0.90) 6.86 144 .05 0.05

Note. Standard deviations are in parentheses.

hp 2

686 The Journal of Social Psychology

subscales. Looking at the effects of the type of individuation for each ethnicity separately, we found that while Euro-Americans expressed equal willingness to engage in either type of behavior, F(1,68) = 1.40, p = .24, hp

2 = .02, Asian Americans expressed more willingness to engage in Taking the Lead actions than Seeking Attention ones, F(1,75) = 12.54, p < .01, hp

2 = .14. We turned next to the scenarios. First, an index of both comfort engaging in

and personal acceptance of the scenario behaviors was computed by calculating the mean of the six scenarios (as were .64 and .57, respectively). Euro-Americans expressed more comfort than Asian Americas engaging in the individuating behavior (there was also a main effect of sex, F[1,144] = 6.19, p < .05, hp

2 = .04, such that men [M = 2.93] indicated more comfort than women [M = 2.59], but no ethnicity by sex interaction). Euro-Americans also reported more acceptance of the scenario behaviors.

The role of social appropriateness. Next, we examined the role of judgments of social appropriateness in comfort engaging in the individuating behavior. Both Euro- and Asian Americans seemed to consider the social appropriateness of the behavior when deciding whether they would do it or not, as the correlation between social appropriateness and comfort was large for both groups, r(68) = .45 and r(75) = .52, respectively, both ps < .001. To examine whether these correla- tions were significantly different from each other, a hierarchical regression was performed. Nominal variables were dummy coded (Euro-American = 0, Asian American = 1), variables were standardized, and interaction terms were com- puted with the standardized values (Cronbach, 1987). This revealed effects of ethnicity, b = −.23 (p < .01), and social appropriateness, b = .51 (p < .001; there was also an effect of sex, b = −.21 [p < .01], but no interaction with sex). The interaction of ethnicity and social appropriateness was not significant, b = −.01, p = .88. Thus, the relationship between judgments of normativeness and comfort engaging in individuating behavior was not moderated by ethnicity.

For acceptance of individuating behavior, both groups again seemed to consider social appropriateness, as the correlation between social appropriateness and acceptance was marginal for Euro-Americans, r(68) = .23, p < .06, and significant for Asian Americans, r(75) = .64, p < .001. An identical hierarchical regression was performed, which revealed effects of ethnicity, b = −.27 (p < .001), social appropriateness, b = .44 (p < .001), and the predicted ethnicity by social appropriateness interaction, b = .21 (p < .01). Thus, the relationship between judgments of normativeness and personal acceptance of individuating behavior was moderated by ethnicity, in that considerations of social appropriateness were more highly related to acceptance for Asian Americans than Euro-Americans.

The role of self-construals. First, we ran mediation analyses using the self- construal subscales, and then re-ran them using the autonomy and group esteem facets. To examine ethnic differences in the subscales, we performed a 2 (ethnicity: Euro-, Asian American) × 2 (self-construal subscale: independent, interdepen- dent), mixed-model ANOVA, which revealed a marginal effect of subscale,

Boucher & Maslach 687

F(1,143) = 3.16, p = .08, hp 2 = .02, such that interdependent items (M = 4.78)

were more highly endorsed than independent ones (M = 4.63), and a significant ethnicity by subscale interaction, F(1,143) = 17.52, p < .001, hp

2 = .11. We decomposed this interaction and found that Euro-Americans scored more highly on independence than Asian Americans, while Asian Americans scored more highly on interdependence than Euro-Americans (see Table 1). The subscales were uncorrelated, r(144) = −.02, p = .80.

To test for the mediating effect of the subscales on both willingness to indi- viduate oneself and acceptance of others’ individuating behavior, we followed Baron & Kenny’s (1986) recommendations. Both ethnicity and the independence subscale were associated with the Individuation Scale (bs = −.36 and .64, respec- tively, both ps < .001), Taking the Lead (b = −.26, p < .01, and b = .56, p < .001), Seeking Attention (bs = −.41 and .61, both ps < .001), comfort engaging in the behaviors from the scenarios (b = −.21, p < .05, and b = .36, p < .001), and accep- tance of them (b = −.21, p < .05, and b = .27, p < .01). Ethnicity was also associ- ated with independence, b = −.17 (p < .05). When we controlled for independence, ethnicity remained significant for the full Individuation Scale (b = −.26, p < .001), Taking the Lead (b = −.17, p < .05), and Seeking Attention (b = −.31, p < .001). However, the drops in beta (.10, .09, and .10, respectively) were significant in all cases (all ps < .05), indicating partial mediation. The effect of ethnicity became marginal on comfort engaging in the scenario behaviors (b = −.15, p = .06), and remained significant for acceptance (b = −.18, p < .05). However, both drops in beta (.06 and .03, respectively) were marginal (ps = .07 and .10), suggesting no mediation. Interdependence was not associated with the Individuation Scale, b = −.14 (p = .11), or Taking the Lead, b = −.01 (p = .89), but was associated with Seeking Attention, b = −.24 (p < .01), and both comfort engaging in and acceptance of the scenario behaviors, bs = −.21 and −.20, both ps < .05. Ethnicity was also associated with interdependence, b = .31 (p < .001). When controlling for ethnicity, the effect of ethnicity remained significant for both Seeking Attention (b = −.37, p < .01) and acceptance (b = −.17, p < .05), and became marginal for comfort (b = −.16, p = .07), but the drops in beta (.04, .04, and .05, respectively) were not significant, p = .13, .12, and .09, indicating no evidence for mediation by interdependence. Thus, while independence partially mediated ethnic differences in the full Individuation Scale and both subscales, independence failed to mediate ethnic differences in both comfort engaging in and acceptance of the scenario behaviors, and interdependence failed to mediate ethnic differences in any of the dependent variables.

Next, we turned to the specific facets of self-construals. We performed a 2 (ethnicity: Euro-, Asian American) x 2 (facet: autonomy, group esteem), mixed- model ANOVA, which revealed an effect of facet, F(1,143) = 69.75, p < .001, hp

2 = .33, such that scores were higher on group esteem (M = 5.07) than auton- omy (M = 4.26), and a significant ethnicity by facet interaction, F(1,143) = 25.71, p < .001, hp

2 = .15. We decomposed this interaction and found that Euro-Americans

688 The Journal of Social Psychology

scored more highly than Asian Americans on autonomy while Asian Americans scored more highly than Euro-Americans on group esteem (see Table 1). The facets were uncorrelated, r(144) = −.06, p = .48.

We present results of the mediation analysis for the specific facets of self- construal in Table 2. Both ethnicity and autonomy were significantly associated with Taking the Lead, and comfort engaging in and acceptance of the scenario behaviors (see Table 2), and ethnicity was significantly associated with auton- omy, b = −.33, p < .001. When we controlled for autonomy, the effect of ethnic- ity on Taking the Lead, comfort, and acceptance became non-significant. Sobel’s test indicated a significant drop in beta, suggesting full mediation of these depen- dent variables. While the effect of ethnicity on Seeking Attention and the full Individuation Scale remained significant when controlling for autonomy, Sobel’s test indicated a significant drop in beta, indicating partial mediation.

Note that earlier we found that ethnicity moderated the relationship between judgments of social appropriateness and personal acceptance of individuating behavior, such that Asian Americans were more likely to express acceptance to the extent that the actions were perceived to be socially appropriate. We also tested for mediated moderation; that is, whether autonomy mediated this modera- tion effect. Autonomy marginally predicted judgments of social appropriateness, b = .14 (p = .09), and when we controlled for autonomy, the interaction of ethnic- ity and social appropriateness became marginal, b = .13 (p = .09). Despite this,

TABLE 2. Summary of Mediational Analyses for Autonomy Facet of Self-Construal Scale

Predictor

Ethnicity Autonomy Ethnicity/autonomy Sobel’s test

Dependent Measure b p b p b p change in b p

Individuation Scale

−.36 .001 .69 .001 −.16 .05 .20 .001

Taking the Lead

−.26 .01 .58 .001 −.08 .28 .18 .001

Seeking Attention

−.41 .001 .67 .001 −.21 .01 .20 .001

Individuating Scenarios Comfort −.18 .05 .48 .001 −.06 .47 .12 .001 Acceptance −.22 .01 .29 .001 −.14 .11 .08 .05

Note. Ethnicity/autonomy is the effect of ethnicity after controlling for autonomy scores.

Boucher & Maslach 689

Sobel’s test indicated that this drop in b (.08) was marginal, p = .11. Thus, ethnic differences in autonomy did not explain this moderation effect.

Next, we re-ran all of the above analyses, substituting group esteem for autonomy. Group esteem was not significantly related to the full Individuation Scale (b = −.04, p = .66), Taking the Lead (b = .06, p = .51), or Seeking Attention (b = −.13, p = .13). While group esteem was significantly related to both comfort engaging in and acceptance of the scenario behaviors, bs = −.20 and −.18, respec- tively, both ps < .05, the effect of ethnicity remained significant when we controlled for group esteem, bs = −.17 and −.18, respectively, both ps < .05 (change in beta = .04 and .03, p = .11 and .16). Thus, ethnic differences in indi- viduation were not mediated by group esteem.

DISCUSSION

The present study provides fairly strong evidence for the relationship between culture and individuation. Not only did we replicate ethnic differences in willing- ness to individuate oneself, but we also provided initial evidence for why these dif- ferences exist. Differences in autonomy/assertiveness between Euro-Americans and Asian Americans fully mediated the effect of ethnicity on willingness to engage in the Taking the Lead behaviors in the Individuation Scale and both comfort engaging in and personal acceptance of the behavior in the scenarios, and partially mediated differences in the Seeking Attention subscale of the Individua- tion Scale. At the same time, while Asian Americans scored more highly than Euro-Americans on both, neither interdependence as a whole nor the group esteem facet mediated any of our ethnic differences. While some group esteem items seem to be conceptually related to a desire not to individuate oneself (e.g., desiring to maintain group harmony, having respect for authority figures), others do not (e.g., considering parental advice before making important decisions, offering one’s seat on the bus to a professor), which could explain this null finding.

It is also interesting to note the role that norms play in individuation. As we anticipated, Euro-Americans were equally likely to express comfort engaging in Taking the Lead and Seeking Attention individuating actions, while Asian Americans were significantly more comfortable engaging in Taking the Lead ones. To the extent that Taking the Lead actions are more socially appropriate than Seeking Attention ones, this provides indirect evidence that Asian Americans are more attuned to the social appropriateness of individuating behavior. More direct was evidence that Asian Americans were more likely than Euro-Americans to consider the social appropriateness of the scenario behaviors when rating their personal acceptance of them. However, this difference could not be accounted for by self-construals (or by any of the facets identified by Hardin et al., 2004). Self- construal is a broad construct, and theoretically the role of norms in determining behavior is included in it (Markus & Kitayama, 1991). However, the Self-Construal Scale does not contain items measuring this aspect of the construct (or other

690 The Journal of Social Psychology

aspects, such as the role of others in social comparison; Hardin et al., 2004). This may explain why self-construals failed to explain the relationship between social appropriateness and reactions to the behavior in the scenarios. It is also important to state that Euro-Americans were affected by normative considerations as well, significantly where comfort engaging in individuating behavior was concerned and marginally in the case of personal acceptance of it. It is simply not the case that Euro-Americans ignore norms; indeed, research on conformity pays tribute to just how susceptible to norms they can be. Moreover, it is likely that achieving distinctiveness that is nevertheless socially approved is one strategy people enact in order to achieve the level of moderate distinctiveness that is most desired according to uniqueness theory (Lynn & Snyder, 2002); that is, people can appear different, fulfilling their need for distinctiveness, while also living up to social norms, fulfilling their need for similarity to others.

The current findings do have limitations. The self-report nature of the study makes it difficult to determine whether participants would individuate them- selves to the extent to which they say they would. The Euro-Americans may have exaggerated their willingness to individuate themselves in order to look good, especially given the fact that conformity has negative connotations in the U.S. (Kim & Markus, 1999). It may be that they simply wish to appear unique. This concern is mitigated somewhat by the already-demonstrated predictive validity of the Individuation Scale (Maslach et al., 1985, 1987). However, a behavioral study is needed to truly ascertain the extent to which ethnic differences in indi- viduation exist in the real world.

Despite this limitation, this research has important implications, especially in the fields of education and health. Many courses in the U.S. require students to participate in class discussions as part of their final grade. We suggest that Asian American students may not be as comfortable doing this as their Euro-American counterparts, because they are less likely to want to assert themselves and see class participation as tantamount to showing off (Paulhus, Duncan, & Yik, 2002). One possible solution is to have students participate in semi-structured class activities (e.g., group presentations); anecdotally, Asian Americans perform just as well as Euro-Americans in this kind of format (C. Maslach, personal commu- nication, October 26, 2000). In addition, cultural differences in individuation may help explain the tendency for Asian Americans to fail to report mental health problems (Tracey, Leong, & Glidden, 1986). In addition to reasons offered by previous researchers for this phenomenon (e.g., saving face; Tracey et al., 1986), it may also be that Asian Americans want to avoid the spotlight that can follow from seeking help for emotional or interpersonal problems. It may be the lesser of two evils for Asian Americans to suffer in silence than to deal with the repercussions of standing out that comes from admitting to difficulties.

An interesting direction for future research is biculturalism; that is, cultural identities coexisting in one person. There is increasing evidence that the process of acculturation does not involve the replacement of one culture with another, but

Boucher & Maslach 691

rather the retention and internalization of both cultures (Ryder, Alden, & Paulhus, 2000). One or the other culture then becomes salient depending on various fac- tors, such as cues in the environment (Hong et al., 2000). The Asian Americans used in the current research likely have internalized two cultures (since the majority of them are “second generation”), and not individuating themselves may be the default response based on their Asian heritage. It would be interesting to examine whether these participants individuate themselves when other cues are made salient, such as their identity as Americans. Future research should address this possibility.

Finally, future research should address the kinds of individuation that may be more acceptable for members of cultures where assertiveness and autonomy strivings are less prevalent. Tafarodi et al. (2004) made a distinction between horizontal distinctiveness, manifested through unusual preferences, hobbies, clothing, and the like, and vertical distinctiveness, which is tied to success or failure in some domain. They suggest that members of collectivistic cultures may shy away from the former kinds of distinctiveness, as they connote the flaunting of social convention. Vertical distinctiveness, however, as it is earned through the pursuit of excellence in socially-sanctioned domains may be perfectly accept- able. Similarly, Vignoles and his colleagues (2000) made distinctions among difference, or distinctiveness achieved via one’s internal characteristics, such as abilities and opinions, separateness, referring to independence or distance from others, and achieved by physical separation from others or more subjective feelings of uniqueness, and position, or distinctiveness tied to one’s place within social relationships (e.g., one’s unique role in one’s family). Vignoles et al. (2000) suggest that this last kind of distinctiveness may receive greater emphasis in collectivistic cultures. The horizontal distinctiveness of Tafarodi et al. (2004) seems to map well onto both difference and separateness (and Seeking Atten- tion), while vertical distinctiveness seems to map onto distinctiveness achieved via position in Vignoles et al.’s (2000) model (and Taking the Lead). Future research should examine the relationships among these constructs, and identify new ones.

Cross-cultural researchers have been criticized for documenting “mere” ethnic differences, without specifying what it is about ethnicity that is impacting the relevant dependent variable (Matsumoto, 1999). Partially in response to these criticisms, researchers offered dimensions of cultural variability such as individ- ualism/collectivism (Triandis, 1995), independent/interdependent self-construals (Markus & Kitayama, 1991), and analytical/holistic folk epistemologies (Peng & Nisbett, 1999). Unfortunately, the dominant trend since these theoretical advances is to specify the construct causing ethnic/national differences, find the predicted differences, and then infer that the differences are due to the con- struct, without measuring it (Matsumoto, 1999). The present research is thus a demonstration of the utility of specifying the construct(s) believed to underlie ethnic/national variation, measuring them, and finding that they do mediate or

692 The Journal of Social Psychology

moderate the relationship between ethnicity and other variables of interest. We hope we have also demonstrated the importance of specifying the dimension of self-construal most pertinent to particular dependent variables. The autonomy facet of independence was a more powerful mediator of ethnic differences in individuation than independence as a whole. This adds to other work showing that specific facets of self-construal explain more variance than whole sub- scales (Hardin, 2006). In sum, future research should examine cultural differ- ences in various kinds of distinctiveness more systematically, while still including measurements of the constructs thought to moderate or mediate any obtained differences.

NOTES

1. All analyses were re-run with both primary language and length of stay in the U.S. as covariates; all significant results remain significant.

2. Hardin et al. (2004) derived their facets from a 30-item version of the Self- Construal Scale. Since the present data was collected before that paper was published, we computed facets from the 24-item scale. Other alphas were as follows: individualism (a = .53), behavioral consistency (.70), primacy of self (.27), relational interdependence (.54).

AUTHOR NOTES

Helen C. Boucher is an assistant professor in the Department of Psychology, Bates College, Lewiston, Maine. Christina Maslach is a professor at the University of California, Berkley.

REFERENCES

Baron, B. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.

Bond, R., & Smith, P. B. (1996). Culture and conformity: A meta-analysis of studies using Asch’s (1952b, 1956) line judgment task. Psychological Bulletin, 119, 111–137.

Cronbach, L. (1987). Statistical tests for moderator variables: Flaws in analyses recently proposed. Psychological Bulletin, 102, 414–417.

Hardin, E. B. (2006). Convergent evidence for the multidimensionality of self-construal. Journal of Cross-Cultural Psychology, 37, 516–521.

Hardin, E. B., Leong, F. T. L., & Bhagwat, A. A. (2004). Factor structure of the Self-Construal Scale revisited: Implications for the multidimensionality of self-construal. Journal of Cross- Cultural Psychology, 35, 327–345.

Heine, S. J., Lehman, D. R., Peng, K., & Greenholtz, J. (2002). What’s wrong with cross cultural comparisons of subjective Likert scales?: The reference-group effect. Journal of Personality and Social Psychology, 82, 903–918.

Hong, Y.-Y., Morris, M. W., Chui, C.-Y., & Benet-Martínez, V. (2000). Multicultural minds: A dynamic constructivist approach to culture and cognition. American Psychol- ogist, 55, 709–720.

Kim, H., & Markus, H. R. (1999). Deviance or uniqueness, harmony or conformity? A cultural analysis. Journal of Personality and Social Psychology, 77, 785–800.

Boucher & Maslach 693

Kwan, V. S. Y., Bond, M. H., Boucher, H. C., Maslach, C., & Gan, Y.-Q. (2000). Individ- uation: More complex in collectivistic than individualistic cultures. Personality and Social Psychology Bulletin, 28, 300–310.

Lam, A. G., & Zane, N. W. S. (2004). Ethnic differences in coping with interpersonal stressors: A test of self-construals as cultural mediators. Journal of Cross-Cultural Psychology, 35, 446–459.

Lynn, M., & Snyder, C. R. (2001). Uniqueness seeking. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 395–410). New York, NY: Oxford University Press.

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253.

Maslach, C. (1974). Social and personal bases of individuation. Journal of Personality and Social Psychology, 29, 411–425.

Maslach, C., Santee, R. T., & Wade, C. (1987). Individuation, gender role, and dissent: Personality mediators of situational forces. Journal of Personality and Social Psychol- ogy, 53, 1088–1093.

Maslach, C. Stapp, J., & Santee, R. T. (1985). Individuation: Conceptual analysis and assessment. Journal of Personality and Social Psychology, 49, 729–738.

Matsumoto, D. (1999). Culture and self: An empirical assessment of Markus and Kitayama’s theory of independent and interdependent self-construals. Asian Journal of Social Psychology, 2, 289–310.

Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128, 3–72.

Paulhus, D. L., Duncan, J. H., & Yik, M. S. M. (2002). Patterns of shyness in East-Asian and European-heritage students. Journal of Research in Personality, 36, 442–462.

Peng, K., & Nisbett, R. E. (1999). Culture, dialectics, and reasoning about contradiction. American Psychologist, 54, 741–754.

Ryder, A. G., Alden, L. E., & Paulhus, D. L. (2000). Is acculturation unidimensional or bidimensional? A head-to-head comparison in the prediction of personality, self-identity, and adjustment. Journal of Personality and Social Psychology, 79, 49–65.

Singelis, T. M. (1994). The measurement of independent and interdependent self-construals. Personality and Social Psychology Bulletin, 20, 580–591.

Tafarodi, R. W., Marshall, T. C., & Katsura, H. (2004). Standing out in Canada and Japan. Journal of Personality, 72, 785–814.

Tracey, T. L., Leong, F. T. L., & Glidden, C. (1986). Help seeking and problem perception among Asian Americans. Journal of Counseling Psychology, 33, 331–336.

Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press. Vignoles, V. L., Chryssochoou, X., & Breakwell, G. M. (2000). The distinctiveness

principle: Identity, meaning, and the bounds of cultural relativity. Personality and Social Psychology Review, 4, 337–354.

Whitney, K., Sagrestano, L. M., & Maslach, C. (1994). Establishing the social impact of individuation. Journal of Personality and Social Psychology, 66, 1140–1153.

Yamaguchi, S., Kuhlman, D. M., & Sugimori, S. (1995). Personality correlates of allocentric tendencies in individualistic and collectivistic cultures. Journal of Cross-Cultural Psychology, 26, 13–25.

Ybarra, O., & Trafimow, D. (1998). How priming the private self and or collective self affects the relative weights of attitudes and subjective norms. Personality and Social Psychology Bulletin, 24, 362–370.

Received June 15, 2008 Accepted September 23, 2008

Copyright of Journal of Social Psychology is the property of Taylor & Francis Ltd. and its content may not be

copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

Copyright of Journal of Social Psychology is the property of Taylor & Francis Ltd. and its content may not be

copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.