We sought to investigate whether there are differences in self-esteem among individuals who differ in their exposure to Western culture. We had included the Rosenberg Self-Esteem Scale (Rosenberg, 1965) in a large number of questionnaire studies that were conducted with students from universities in Vancouver, Canada and in a variety of cities in Japan. We created a large file that included participants’ self-esteem responses and some demographic variables (a total of over 5000 participants). The participants came from the University of British Columbia and Simon Fraser University in Canada, and from Aichi Gakuin, Doshisha University, Kansai Gaikokugo UniversityUniversity, Kyoto University, Nagasaki University, Nara University, Ritsumeikan University, and Toyama University in Japan. Japanese participants completed the scale in Japanese and Canadian participants completed the scale in English. The original Rosenberg Scale was translated into Japanese, back-translated into English, and any discrepancies between the two versions were discussed among three translators.
As a large proportion of university students in the two Canadian universities are of Asian descent, from a variety of different countries with the most common ethnic heritage being Chinese (self-criticism is also evident among Chinese; e.g., Yik, Bond, & Paulhus, 1998), and that a significant number of the Japanese students had spent time in a Western country, we were able to analyze the data with respect tolyze the data with respect to how much time participants had been exposed to Western culture2. A continuum of increasing exposure to Western culture was created by classifying participants into the following groups: 1) Japanese who had never been outside of Japan (n=1657); 2) Japanese who had spent some time in a Western country (n=577); 3) Recent Asian immigrants to Canada (n=244); 4) Long-term Asian immigrants to Canada (n=289); 5) Second generation Asian-Canadians (n=431); 6) Third generation Asian-Canadians (n=38); 7) European-Canadians (n=1466). A total of 388 participants from a variety of ethnic backgrounds did not fit any of these categories and were not included in the analyses.
We conducted a culture by sex ANOVA on self-esteem for the entire sample. A pronounced difference for culture emerged, F(6, 4690) = 244.86, p < .001, which is depicted in Figure 1. Replicating past research, European-Canadians scored higher on self-esteem than did Japanese (they scored higher on 9 of the 10 items; the item, “I certainly feel useless at times” showed no cultural difference). The other cultural groups formed a remarkably monotonically increasing pattern between these two extremes. Self-esteem rose among people of Asian descent with exposure to Western culture to the point that 3rd-generation Asian-Canadians had self-esteem scores that approximated those of European-Canadians. The more exposure individuals had to cultural situations, scripts, and institutions associated with higher self-esteem, the more positively they viewed themselves. The small size of the 3rd-generation Asian-Canadian sample warrant caution in interpreting the results, but if we assume it is reliable this suggests that three generations is enough for people of Asian descent to fully acculturate to Canadian culture in terms of their self-esteem3.
Insert Figure 1 about here.
This cCross-sectional studies such as this do have some interpretative limitations. For example, there may be cohort effects distinguishing the different cultural groups in terms of a number of demographic variables, such as their reasons for migrating to Canada, their past education history, or their performance at school, which may relate to their self-esteem scores. We felt it was imperative to replicate this basic finding employing a controlled longitudinal design in order to avoid these interpretive ambiguities. Three separate longitudinal studies were conducted in which individuals’ self-esteem was measured at 2 points in time: 1) before leaving one’s home culture (or just after arriving in the host culture) and 2) 7 months after having lived in the host culture.
Method and Results
Two days after arriving in Vancouver to begin an 8-month exchange program, 84 students from Ritsumeikan University (out of 99 who were enrolled in the program) agreed to participate in a questionnaire study. One of the measures in the questionnaire was Rosenberg’s (1965) Self-Esteem Scale. Approximately 7 months later, in one of their classes, the students were invited to attend an evening lecture during which a second questionnaire, which also included the Rosenberg Scale, was distributed. Participants completed Japanese versions of the scale at both points in time.
Unfortunately, because many students were not in class when the announcement was made, only 35 students attended the lecture and participated in Wave 2 of the study4. A repeated measures ANOVA was conducted on the 35 students who participated in both waves of the questionnaire. Subjects Participants showed higher self-esteem at Wave 2, M = 38.5, than at Wave 1, M = 36.7; F[1, 33] = 4.04, p = .052 (see Figure 2). Hence, Japanese exchange students exhibited an increase in their self-esteem after living in Canada for 7 months.
Insert Figure 2 about here.
A potential confound of Study 2a is that acculturation experiences per se might have led to the self-esteem increases of the Japanese sample. Perhaps anyone who moves to a new cultural environment, regardless of their cultural background or destination, experiences increases in self-esteem due to their expanding horizons and feelings of competence associated with being able to survive in a foreign environment. To the extent that it is experiences in North American culture that are associated with self-enhancement (or experiences in Japanese culture that are associated with self-criticism; Heine et al., 1999; Kitayama et al., 1997), and not acculturation experiences per se, we should expect to see self-esteem decreases among North Americans moving to Japan. Study 2b investigated this possibility.
Method and Results
Shortly before leaving Canada, 73 Canadian English teachers who were heading to Japan to participate in the Japan Exchange and Teaching (JET) program completed a questionnaire packet including Rosenberg’s Self-Esteem Scale. Seven months later the teachers were mailed a second questionnaire that also included the Rosenberg Scale. Sixty-nine of the teachers completed the second wave of the study. A repeated measures ANOVA reveals that the teacher’s’ self-esteem was lower at Wave 2, M = 42.16, than at Wave 1, M = 43.15; F (1, 67) = 4.93, p < .03. Canadian English teachers thus displayed a decrease in their self-esteem after living in Japan for 7 months. Acculturation experiences per se are not associated with increasing levels of self-esteem. Canadians who were removed from a cultural environment that bolsters self-esteem and placed in an environment characterized by various practices associated with self-criticism (e.g., Heine et al., 1999; Lewis, 1995) appeared to become more self-critical after 7 months.
Study 2c sought to replicate the self-esteem increase among Japanese students in Study 2a, and to explore whether acculturation attitudes moderated the relation between mere exposure to Western culture and the internalization of Western cultural norms.
Method and Results
One month prior to leaving Japan, 82 Ritsumeikan University students (out of 93 who were enrolled in the program) who were heading to Vancouver as exchange students were given a questionnaire packet including Rosenberg’s Self-Esteem Scale. Seven months after arriving in Canada, 74 students completed a second questionnaire as part of a class project, which also included Rosenberg’s Self-Esteem Scale as well as John Berry and colleagues’ Acculturation Attitudes Scale (Berry, Kim, Power, Young, & Bujaki, 1989), which was modified specifically for Japanese exchange students coming to Canada (Davis, 1995). This scale assesses the positivity of students’ attitudes towards Canada and Japan, and this scale was included to investigate whether individual differences in attitudes towards Canada relate to self-esteem change.
A repeated measures ANOVA revealed that students’ self-esteem scores were nominally, although not significantly, higher at Wave 2 (M = 34.5) than at Wave 1 (M = 34.2; F < 1). Hence, we failed to replicate the significant increase in self-esteem among Japanese exchange students living in Canada that was demonstrated in Study 1.
An overall composite of subjects’ acculturation attitudes was formed by summing all the items expressing positive attitudes towards Canada (Assimilation and Integration subscales) and subtracting the items expressing negative attitudes towards Canada (Separation and Marginal subscales). This total value reflects how positive students’ attitudes were towards Canadian culture, and is a proxy for how much students made efforts to “become Canadian” while on the exchange program. This total acculturation score was then correlated with participants’ self-esteem change scores. This analysis revealed a positive relation between how positive participants’ attitudes were towards Canadian lifestyles and how much their self-esteem increased, r = .32, p < .01. That is, the more participants were open to Canadian culture (and theoretically the more they were influenced by Canadian cultural values), the more their self-esteem increased during their stay in Canada. This provides another source of evidence to suggest that the self-concepts of those who were participating in Canadian culture were influenced by the cultural environment.
The investigation of the ways in which culture affects the self-concept is fraught with methodological obstacles. For example, there are no appropriate control groups of “cultureless” humans with which to compare the different varieties of “cultured” ones, there are no direct measures of cultural grammars, and people cannot be randomly assigned to different cultural environments. One quasi-experimental approach, however, which rarely has been pursued, investigates changes in the self-concept that occur during sojourns to new cultures.
We investigated acculturative effects on an evaluative component of the self-concept: global self-esteem. Much research has maintained that self-esteem, as it has traditionally been operationalized within Western psychology, is a construct that is enhanced by participation in North American culture (e.g., Heine, in press-a; Heine et al., 1999; Kitayama et al., 1997; Markus & Kitayama, 1991). The results of the present studies provide further evidence that self-esteem is intimately related with Western cultural values. Study 1 revealed a clear relation between self-esteem and exposure to North American culture in a large-scale cross-sectional study. Study 2a demonstrated that time spent in Canada led to an increase in self-esteem for Japanese students, whereas Study 2b revealed that time spent in Japan led to a decrease in self-esteem for Canadians. Exposure to new cultural environments seems to have been associated with movement in sojourners’ self-esteem towards levels that are normative of their host cultures. Study 2c failed to replicate the significant self-esteem increase found in Study 2a, but demonstrated that those Japanese most receptive to Canadian cultural values displayed a greater increase in self-esteem than those who resisted the host culture. This relation is also consistent with the notion that greater exposure to Western culture leads to higher self-esteem. Taken together, these four studies are suggestive of a significant Western cultural component in the construct of self-esteem. One interpretation of these results, consistent with past research with biculturals (e.g., Hong et al., in press2000), is that the longer one is in a culture the more likely it is that the metaschema of thoughts and feelings that is activated in them is associated with the host culture.
Cultural differences in self-esteem between Japanese and North Americans appear to yield some of the largest effect sizes between cultures of any of the aspects of the self-concept investigated thus far (the effect size in Study 1 between the European-Canadians and Japanese who had never been abroad was 1.35), and there is much theoretical reasoning consistent with these cultural differences (e.g., Heine, in press-a; Heine et al., in press; Heine et al., 1999; Kitayama et al., 1997; Markus & Kitayama, 1991). In these respects, self-esteem is an especially useful tool for identifying acculturative effects on the self-concept. However, self-esteem is likely also confounded by experiences of success and failure that are part and parcel of the acculturation experience, and thus the self-esteem assessments obtained here are unlikely to be pure measures of acculturation. Future research investigating other aspects of the self-concept known to be influenced by culture, such as tendencies to make situational attributions (Hong et al., 2000), perceptions of agency (Morling, Kitayama, & Miyamoto, in press), self-improving motivations (Heine et al., in press) or feelings of independence (Singelis, 1994) could move the field forward.
In the present studies we compared mean scores on subjective Likert scale attitude measures across cultural groups. Such comparisons can potentially be undermined by reference-group effects (Heine et al., 2001). That is, people evaluate themselves by implicitly comparing themselves to those around them. What makes this problematic for cross-cultural comparisons is that people from different cultures are comparing themselves to different referents (Biernat & Manis, 1994; Heine et al., 2001; Heine et al., in press; Peng et al., 1997). However, that we found evidence for Japanese self-esteem increasing when surrounded by higher self-esteem Canadians, and for Canadian self-esteem decreasing when surrounded by lower self-esteem Japanese is not consistent with the notion that the findings are due to the different reference groups of the samples.
Much past research on acculturation has assumed that the lower self-esteem scores among sojourners and immigrants reflects the psychological distress inherent in difficulties in the acculturation experience (e.g., Furnham & Bochner, 1986; Taft, 1977). Although it is possible that negative experiences associated with “culture shock” lead to lower self-esteem, the present findings suggest that this is not the best explanation to account for the relatively low self-esteem of immigrating Asians. Japanese have the lowest self-esteem scores before they have left their country, and the self-esteem scores of Asian immigrants to in North America are only low relative to those of European-descent North Americans or 2nd and 3rd generation Asian-descent North Americans. In comparison to their compatriots in their home cultures, the self-esteem scores of Asian sojourners and immigrants are relatively high.
That we observed changes in self-esteem in sojourns as brief as 7 months provides testimony to the influences culture has on the self-concept. Moreover, that these differences were found with young adults suggests that people continue to seek cultural meaning systems even after they have been socialized in a different culture. To the extent that there is a sensitive period for acquiring a cultural meaning system before puberty (e.g., Minoura, 1992), we assume that the changes in self-esteem observed in the present studies would have been larger had we conducted the study with prepubescent children. The clearest evidence of acculturative effects on the self-concept should be observable among children, whose more plastic minds are still adjusting to the cultural meanings with which they interact.