Evidence for Self-Criticism Among Japanese
Steven J. Heine
University of Pennsylvania
Darrin R. Lehman
University of British Columbia
Please address correspondence to:
Steven J. Heine
Department of Psychology
3815 Walnut Street
University of Pennsylvania
Philadelphia, PA 19104
Revision submitted to Personality and Social Psychology Bulletin.
RUNNING HEAD: JAPANESE SELF-CRITICISM
Although a robust finding in cross-cultural research is that Japanese exhibit less self-enhancement than North Americans, all of these studies have employed questionnaire measures susceptible to self-presentational biases. The present study assessed self-enhancement in a laboratory that covertly measured participants’ behaviors. Whereas Canadians were reluctant to conclude that they had performed worse than their average classmate, Japanese were hesitant to conclude that they had performed better. This research provides evidence that cultural differences in self-enhancement and self-criticism go beyond mere self-presentation.
Beyond Self-Presentation: Evidence for Self-Criticism Among Japanese
The assumption of a need for positive self-regard has traditionally been viewed as a human universal (e.g., Maslow, 1943; Tesser, 1988), and tendencies to self-enhance have been assumed to reflect this need (e.g., Greenwald, 1980; Taylor & Brown, 1988). However, some cultural psychologists have questioned the universality of this view (e.g., Heine, Lehman, Markus, & Kitayama, 1998). Indeed, the Japanese self seems better characterized by a need to secure a positive view from others (i.e., a need for face), rather than from oneself, and this securing of others’ approval seems better served by self-improvement rather than self-enhancement (Heine, Lehman et al., 1998; Kitayama, Markus, Matsumoto, & Norasakunnkit, 1997). Self-improvement involves the individual being keenly sensitive to his or her shortcomings that may potentially jeopardize the favorable views of significant others. The individual attends to these self-perceived deficiencies and focuses his or her efforts on correcting them, thereby conveying to others that one is making efforts to become the best that one can. This theoretical framework of self-evaluations predicts that Japanese should exhibit less evidence of self-enhancement than North Americans, and even evidence of self-criticism.
A large number of studies have found such cultural differences. Although there is a great deal of within-culture variability, and the distributions from the two cultures overlap considerably, pronounced cultural differences in self-enhancement have emerged in a range of different experimental paradigms: attributional self-serving biases (Kitayama, Takagi, & Matsumoto, 1995; Meijer & Semin, 1998); unrealistic optimism biases (Heine & Lehman, 1995a); and false uniqueness biases (Heine & Lehman, 1997a; Markus & Kitayama, 1991a). Moreover, these cultural differences appear robust on a number of fronts: for example, they do not appear to be due to a biased coverage of content, as Japanese are especially self-critical in situations that they generate themselves (Kitayama et al., 1997) and Japanese view themselves as even further away from their ideals than do North Americans for the traits that they view as most important for succeeding in their cultures (Heine & Lehman, in press). These cultural differences are not restricted to evaluations of individuals: cross-cultural comparisons also reveal more pronounced group-serving biases among North Americans compared to Japanese (Heine & Lehman, 1997a).
Are Cultural Differences in Self-Enhancement Merely Due to Differences in Self-Presentation?
Although the cross-cultural literature demonstrates that on average Japanese exhibit less self-enhancement than North Americans, all of this research has been conducted with questionnaire measures. Indeed, the vast majority of research on self-enhancement in North America also has been carried out via questionnaire. This raises the possibility that the observed cultural differences in self-enhancement may be due, at least in part, to cultural differences in how individuals present themselves in questionnaires. For example, it can be argued that North American self-enhancement may be exaggerated because of norms encouraging people to present themselves as confident (cf., Baumeister, Tice, & Hutton, 1989). Self-confidence is an important value in North America (Heine, Lehman et al., 1998), and North Americans may emphasize their self-confidence as a means to earn the approval of others. Whether this hypothesized exaggeration of self-confidence colors North Americans’ responses to anonymous psychological surveys is unclear, although some evidence suggests that it might (Shedler, Mayman, & Manis, 1993).
That an important virtue (espoused by Confucius) characteristic of many Asian cultures, including Japan, is humility or modesty, renders the issue of interpreting at face value questionnaire studies of self-enhancement in Japan a significant concern. Although immodesty is admonished in Japan, it is not clear whether individuals are being sanctioned against having immodest thoughts or for making immodest statements to others. To the extent that the Japanese concern with modesty is only in the domain of actual expression, it is reasonable to be concerned that the relative lack of self-enhancement by Japanese may be due to them simply saying that they are not better than others, even though deep down they may feel quite differently.
There has been considerable discussion regarding whether Japanese are feigning modesty in their questionnaire responses (Heine & Lehman, 1995a, 1995b; Heine, Lehman et al., 1998; Kitayama et al., 1997; Markus & Kitayama, 1991b). Indeed, one characteristic frequently attributed to Japanese is the great distinction they make between their public presentation (tatemae) and their private feelings (honne; Doi, 1986; Lebra, 1976). Research from a number of Asian cultures is consistent with the notion that people’s behavior is viewed more as a function of situational pressures than as the product of dispositional factors (Miller, 1984; Morris & Peng, 1994; Weisz, Rothbaum, & Blackburn, 1984). Many public situations in Japan and throughout Asia require modest self-presentations (Kitayama, Masuda, & Lehman, 1998), and individuals may stand to lose the respect of their peers if they do not present themselves in a modest, self-effacing manner (e.g., Bond, Leung, & Wan, 1982). Thus, situations in which one says one thing (tatemae) while truly feeling something quite different (honne) may be more frequent in Japan than in North America, and this discrepancy between attitudes and behavior may arouse less dissonance among Japanese than it does for Westerners (cf., Heine & Lehman, 1997b; Kashima, Siegal, Tanaka, Kashima, 1992). The distinction between tatemae and honne may be such an integral part of their lives that Japanese could have difficulty expressing their private thoughts, even anonymously in questionnaires.
However, a review of studies with Japanese is not consistent with the notion that Japanese disguise their responses in questionnaires (Heine, Lehman et al., 1998). For example, self-critical tendencies among Japanese are still evident when response anonymity is manipulated and responses are solicited in private (Kitayama, 1998). We do not know of any evidence that supports the notion that Japanese questionnaire responses are less veridical than those of people from other cultures. The most compelling evidence to challenge the feigned modesty account, however, would come from cross-cultural studies that measure self-evaluative tendencies with hidden behavioral measures, thus precluding the possibility for socially-desirable corrective responses. If participants are unaware that what they are doing is meaningful or even observable, then there is no conceivable reason for them to feign modesty. We sought to provide such an assessment of self-enhancement and self-criticism in a cross-cultural laboratory experiment.
We maintain that North American culture tends to encourage individuals to view themselves positively, whereas Japanese culture tends to encourage people to identify their shortcomings such that they can make efforts towards improving themselves. This suggests that North Americans should be particularly accepting of information indicating that they are better than others, and reluctant to believe information that presents them as worse than others. Japanese, in contrast, are not expected to exhibit this preference for self-enhancing information. In fact, given the importance of identifying and working towards eliminating negative information about the self, Japanese may even be more inclined to accept self-relevant information that is negative.
The present research examined the above hypothesis in a study modeled after a similar study by Takata (1987), which in turn, was modeled after a study by Schwartz and Smith (1975). Schwartz and Smith originally were interested to see if people (Americans) had the implicit understanding of a t-test when they made social comparisons about ability. That is, they examined whether individuals’ inferences about their abilities relative to comparison others were influenced by the effect size between, and the variance within, their respective performances. Schwartz and Smith first had participants compare their performance on a reaction time task with the previous participant, one trial at a time, and then decide which person, themselves or the previous participant, had scored better on the task, after viewing as few trials as possible. The results indicated that participants tended to view more trials before making their decision as the effect size was reduced, or as the variance increased. In effect, participants were calculating intuitive t-tests. Schwartz and Smith also observed that participants were significantly more confident in their decisions when their scores were higher than the previous participants’ scores than when they were lower, although they did not differ in the number of trials that they viewed. That is, the participants self-enhanced with respect to their confidence in their decisions.
Takata (1987) attempted a conceptual replication of this study in Japan, making a number of modifications in the design. Interestingly, he identified a pronounced self-critical effect for Japanese. That is, participants viewed significantly more trials before deciding that they had outperformed the previous participant (success condition) than did their counterparts who decided that the previous participant had outperformed them (failure condition). Participants also were significantly less confident in their decisions in the success condition than in the failure condition. In an attempt to rule out the confounding explanation that participants feared they might meet the previous participant, Takata conducted a second study in which the comparison object was a computer programmed to perform like a typical undergraduate student. Participants still exhibited a highly pronounced self-critical tendency. Japanese were thus more easily convinced that they had performed worse than average than they were that they had performed better than average.
Takata’s (1987) findings are highly suggestive of a self-critical orientation of Japanese. However, the absence of a cross-cultural comparison group makes it impossible to rule out the concern that there were artifactual elements in his novel experimental design that produced these results. North Americans may have behaved in a similar way had they been run in the same study, which would argue against cultural explanations. The present study sought to directly whether Japanese and Canadians differ in self-enhancing or self-critical tendencies.
Canadian participants were introductory psychology students enrolled at the University of British Columbia (UBC). Participants were contacted through the participant pool, and because we wanted to compare a Western sample with the Japanese, participants were selected on the basis of having names that seemed to be of European origin. One hundred and thirty-four participants were run through the experiment, but the data of 4 were eliminated because they expressed suspicion regarding the deception, leaving a total of 130 participants (69 females and 61 males) in the final Canadian sample. Japanese participants were introductory psychology students at Nara University in Nara, Japan. One hundred and twenty-nine participants completed the experiment, but the data of 1 were eliminated because of a computer malfunction. None of the Japanese participants expressed any suspicion regarding the deception. This resulted in a total of 128 participants (58 females and 70 males) in the final Japanese sample.
A number of changes were made to Takata’s (1987) original experimental design. The participants were brought into the lab individually to take part in an experiment entitled “Cognition and Judgment,” which they were told consisted of two phases. Phase 1 was called the “Integrative Cognitive Capacity” (ICC) test, which participants were told was a new kind of intelligence test. Participants viewed grids of colored shapes that were presented on the computer and were asked to answer as quickly and accurately as possible questions regarding the numbers of certain shapes presented. Participants were told that their speed and accuracy in answering these questions reflected how well they could integrate and manipulate information, a skill that they were told is critically linked to intelligence. An example question was “How many more yellow stars are there than red circles?” There were 20 such questions in the ICC test.
Phase 2 of the study was called the “Judgment Under Uncertainty” task. Participants were told that we were interested in how well they could make a mathematical judgment with only a limited amount of information. They were urged to make the judgment as soon as they felt that they had seen enough information to be able to discern the correct answer. The judgment that they were to make was whether their performance on the ICC test was better or worse than that of a computer, which participants were told was programmed to function the same as the average UBC /Nara University student. They were presented with what they were told were their scores for each trial of the ICC test alongside the scores of the average student, one trial at a time (higher scores denote better performance). Participants were told that they had to view at least 5 trials and no more than 20 trials (which was the total number of trials in the ICC test) before indicating that they had enough information to make their decision. The computer randomly assigned participants either to a success or a failure condition, and the experimenter was blind to this assignment. All participants were presented with the same numerical feedback in Phase 2, with the difference between the two conditions being the positioning of the scores: In the success condition, the higher scores were under the Participant’s ID number; in the failure condition, the higher scores were under the Average Student’s column. The feedback was designed such that it would be difficult for participants to make their judgments. The feedback showed a great deal of variability from trial to trial (the standard deviation of the discrepancy between the two columns across all 20 trials was 85.3 points), and the mean difference between the average performance of the participant and the computer was kept small (averaging 18 points across all 20 trials).
After indicating that they had seen enough information to make their decision, participants decided who performed better, themselves or the average student. Next, participants were asked to indicate how confident they were that they had made the correct decision on a scale from 1 (Not at all Confident) to 9 (Extremely Confident). After completing the “Judgment under Uncertainty Phase,” participants completed a brief questionnaire. They were asked to estimate the average ICC scores of both themselves and the average student. This item served as a manipulation check to ensure that participants understood the feedback that they viewed. Then, participants were asked to estimate the percentage of students from their university whom they thought could outperform them on the ICC test, and they were asked whether they believed that the ICC test was an accurate measure of intelligence, on a scale from 1 (Strongly Disagree) to 5 (Strongly Agree). Participants then completed Rosenberg’s (1965) Global Self-Esteem Scale and answered a few demographic questions. The dependent measures in the study were the percentage of errors participants made in their decisions, their confidence in their decisions, the number of trials participants viewed before making their decisions (range 5 to 20 trials), the amount of time participants spent viewing those trials (measured in milliseconds by the computer), their estimates of the percentage of their classmates who could outperform them on the ICC test, and their ratings of the accuracy of the ICC test. After completing the questionnaire participants were probed for suspicion and thoroughly debriefed. All materials and computer instructions were originally produced in English and then back-translated into Japanese by an independent translator.
Comparability of the Samples
The two samples were significantly different with respect to age, F(1, 256) = 8.54, p < .01, with Canadians being older (M = 19.6 years) than Japanese (M = 18.9 years). The samples did not differ in their proportions of females and males, 2(1, N = 258) = 1.56, ns. Analyses were conducted with gender as a factor and age as a covariate, however since none of the conclusions are affected by the inclusion of these variables, we will not discuss them further.
Comparisons were made between participants’ estimates of their own versus the average student’s average ICC scores. This item served as a manipulation check to ensure that participants understood the decision that they were asked to make. Four Canadians and 7 Japanese made estimates that were inconsistent with the decisions they made on the computer (e.g., they estimated their average ICC score to be higher than that of the average student’s, but they had decided that they had been outperformed by the average student). Because these participants did not fully understand the instructions of Phase 2, their results were not included in the final analyses.
Analyses of Self-Enhancement/Self-Criticism
Our attempt at making the mathematical judgment task in Phase 2 challenging to participants turned out to be more successful than we had originally intended. A sizable proportion of the participants (29%) made the incorrect decision regarding whether the trials they viewed indicated that they did either better or worse than the average student. A closer analysis of these errors reveals a noteworthy pattern. Twenty-nine percent of Canadians (21 of 73) assigned to the failure condition mistakenly concluded that they had scored higher than the average student, whereas only 19% (10 of 53) assigned to the success condition incorrectly decided that they were worse than the average student. Hence, Canadians were slightly (but not significantly) more likely to make a self-enhancing decision, concluding that their performance was better than average when the feedback indicated the opposite, than they were to make the incorrect decision in the reverse way. In the Japanese case, 39% (23 of 59) assigned to the success condition incorrectly concluded that they had done worse than average, whereas only 27% (16 of 60) assigned to the failure condition mistakenly concluded that they had in fact done better than average. Japanese thus were slightly (but not significantly) more likely to make a self-critical decision, concluding that they had performed worse than average when the feedback indicated otherwise. The culture by condition interaction of these proportions of errors was marginally significant, 2 (3, N = 245) = 6.95, p < .08, reflecting the tendencies of Canadians to be more reluctant to conclude that they were worse than average and Japanese to be more reluctant to conclude that they were better than average.
The analyses for the remaining dependent variables are complicated by the fact that many participants made errors in their decisions. This is problematic because participants who viewed feedback indicating that they did worse than average, but made incorrect calculations and believed that they did better than average, should respond as though they had received positive feedback. And this would imply that the manipulation yielded the opposite effect that we desired. Another possibility is that participants were so surprised with the feedback of their performance that they didn’t want to believe it, and made a decision opposite to what their calculations of the feedback suggested. Because it is impossible to interpret participants’ reactions to the feedback when they made errors in their decisions, we decided to exclude those participants from the remaining analyses. This leaves a final Canadian sample of 95 (48 females and 47 males) and a Japanese sample of 80 (36 females and 44 males). Although it may seem problematic to exclude such a sizable nonrandom proportion of our sample from analyses, this should be a concern only if it is plausible that the participants that are excluded are more likely to respond in a manner inconsistent with our hypothesis (i.e., if we exclude self-critical Canadians and self-enhancing Japanese). However, as the marginally significant interaction in participants’ errors noted above indicates, Canadians tended to make more self-enhancing errors whereas Japanese tended to make more self-critical errors. Thus, in effect, we are excluding those individuals who are more likely to conform to our hypothesis, thereby rendering the remaining analyses conservative estimates of cultural differences in self-enhancing and self-critical tendencies.
We conducted analyses of how confident participants were that they made the correct decision. An ANOVA revealed a significant culture by condition interaction, F(1, 167) = 9.81, p < .002. Simple effect analyses revealed that Canadians who concluded that they did better than average were marginally more confident that they made the correct decision than those who concluded that they did worse than average, F(1, 91) = 3.70, p < .06 (see Table 1). In contrast, Japanese who concluded that they did worse than average were more confident in their decision than those who concluded that they did better than average, F(1, 76) = 5.63, p < .03. Canadians, therefore, demonstrated a marginally significant self-enhancing tendency, whereas Japanese exhibited significant self-criticism in their decision confidence.
Next we conducted analyses of how many trials participants viewed before making their decision. Again, a significant culture by condition interaction emerged, F(1, 167) = 7.59, p < .007. Simple effect analyses revealed that although Canadians exhibited a slight tendency to view more trials when the feedback indicated that they did worse than average, this effect was not significant, F(1, 91) = 2.60, p = .11. Japanese, in contrast, viewed significantly more trials when the feedback indicated that they did better than average, F(1, 76) = 4.73, p < .04. That is, the Japanese tendency to view more trials before being willing to conclude that they did better than average seems to reflect a self-critical orientation, whereas Canadians exhibited a non-significant self-enhancing tendency.
Analyses of how long participants spent viewing the trials also revealed a significant culture by condition interaction, F(1, 167) = 5.02, p < .03. Canadians viewed the trials significantly longer when they indicated that they did worse than average, F(1, 91) = 5.31, p < .03, whereas Japanese exhibited a non-significant reversal, F(1, 76) = 1.23, ns. Canadians thus were self-enhancing in needing to view the failure trials longer, apparently in disbelief of what they saw, whereas Japanese displayed a hint of self-criticism.
Analyses of the questionnaire data revealed a few cultural differences. First, participants were asked to estimate the percentage of students from their university who could outperform them on the ICC test. An ANOVA revealed a marginally significant culture by condition interaction, F(1, 167) = 3.64, p < .06. Simple effect analyses revealed that regardless of whether Canadians decided that they did better or worse than average, their estimates of the percentage of their classmates that could outperform them remained about the same, F(1, 91) = 1.45, ns. Canadians thought that significantly less than 50% of their classmates could outperform them, regardless of whether they had just viewed feedback indicating that they did better, t(42) = 5.99, p < .001, or worse, t(51) = 5.38, p < .001, than the average student from their school. It seems that Canadians’ self-enhancing tendencies were so strong that they disregarded the highly relevant information about their performance they had just viewed. In contrast, the Japanese estimates were significantly responsive to the feedback that they viewed, F(1, 76) = 18.34, p < .001. Japanese who viewed feedback indicating that they did better than average estimated that significantly less than 50% of their classmates could outperform them, t(35) = 2.97, p < .006. In contrast, Japanese who viewed feedback indicating that they did worse than average estimated that significantly more than 50% of their classmates could do better than them t(43) = 3.19, p < .005. In this respect, Japanese estimates seemed to more accurately reflect the feedback that they were given.
An ANOVA revealed a significant culture by condition interaction with respect to how accurately participants believed the ICC test measured intelligence, F(1, 166) = 4.53, p < .04. None of the simple effect analyses were significant, but we can detect a general trend that Canadians viewed the ICC test as a marginally more accurate measure when they succeeded than when they failed, F(1, 90) = 3.44, p < .07, whereas Japanese tended to view the test as slightly more accurate when they had failed than when they had succeeded, F(1, 76) = 1.38, ns. This same pattern has been observed in other cross-cultural studies of false-feedback between Japanese and Canadians (Heine, Kitayama, Lehman, Takata, & Ide, 1998; Heine & Lehman, 1997b).
Finally, an ANOVA revealed that Canadian participants had significantly higher self-esteem scores (M = 40.4, SD = 6.33) than Japanese (M = 30.2, SD = 5.63), F(1, 169) = 120.2, p < .001. This difference is consistent with past cross-cultural studies (for a review see Heine, Lehman et al., 1998) and is further evidence of the difference in self-enhancing/self-critical tendencies between North Americans and Japanese.
This is the first cross-cultural study to employ behavioral measures of self-enhancement. Taken together, the results provide a compelling case that self-enhancement is more evident among Canadians and self-criticism is more evident among Japanese. Canadians were somewhat reluctant to believe that they had been outperformed by the average student from their university. When their (randomly assigned) feedback implied as much, they exhibited a non-significant tendency to make more errors in their decisions, they were marginally less confident in their decision, they viewed slightly (but not significantly) more information in order to make their decision, they viewed the information for a significantly longer period of time, they still maintained that they could outperform most of their classmates, and they exhibited a marginal tendency to discount the accuracy of the test. These results converge on the notion that Canadians, on average, find it difficult to believe that they were outperformed by their average classmate.
In stark contrast, Japanese were rather reluctant to believe that they could have done better than the average student from their university. It was when they were led to believe that they had performed better that Japanese exhibited a non-significant tendency to make more errors, were significantly less confident in their decision, viewed significantly more information before making their decision, viewed the information for slightly (but not significantly) longer, and displayed a non-significant tendency to discount the accuracy of the test. These results converge on the notion that Japanese find it difficult to believe that they could have outperformed their average classmate. In sum, then, the results point to cultural differences in response to feedback indicating whether participants are above or below average.
As cross-cultural comparisons are inherently correlational in nature (i.e., culture is never manipulated by the experimenter), it is important to consider whether any extraneous factors or third variables may have affected the results. Perhaps the samples differed in ways other than their cultures. The only other characteristics of the samples that were measured in this study, age and gender, did not affect the overall conclusions. Another possibility is that the universities differed in prestige, and although university prestige is difficult to assess across cultures, it would seem fair to say that UBC ranks higher among Canadian universities than Nara University does among Japanese universities. Perhaps the Nara students were not as used to doing well as were the UBC students, and thus were more likely to assume that they did worse than their classmates. However, another recent set of studies that involved students comparing their performance with their classmates with both Nara University students and Kyoto University students (Kyoto University is known as one of the very best schools in Japan) found no difference between the schools in self-critical tendencies (Heine, Kitayama et al., 1998). Hence this does not seem to be a compelling alternative account.
Another potential artifact to consider is that to the extent that intelligence is less of a concern for Japanese than Canadians, Japanese may not care as much whether they are more or less intelligent than average. Stevenson (1995) found that Japanese view achievement to be based more on effort than intelligence, so this does raise the possibility that Japanese are less concerned about their performance on intelligence tests. However, a number of other studies have shown that Japanese are even more self-critical for the traits that they value most (Heine, Kitayama et al., 1998; Heine & Lehman, in press; cf., Kitayama et al., 1997), suggesting that the cultural differences observed here may be even more pronounced if we had selected a trait of greater importance to Japanese.
The present findings are important because past self-enhancement studies were conducted via questionnaire, and the possibility remained that the greater self-enhancement exhibited by North Americans was due to false bravado, and, likewise, the greater self-criticism exhibited by Japanese was owing to feigned modesty. Cross-cultural questionnaire studies are vulnerable to the criticism that they are merely demonstrating cultural differences in response styles and self-presentation. Such criticisms are particularly difficult as they raise the concern that we cannot interpret questionnaire data at face value – a major problem for cross-cultural research if true. Interestingly, the relevant questionnaire data in past cross-cultural studies of self-enhancement are not in support of this self-presentational account (for a review see Heine, Lehman et al., 1998). That Japanese and Canadians’ behavioral responses in the present study were in line with their questionnaire responses, and those in past studies of self-enhancement, strengthens our confidence in the validity of questionnaire measures of self-enhancement, for both cultural groups. Yet it remains important for cross-cultural psychologists to continue exploring new measures and techniques that will enable us to transcend the methodological difficulties inherent in cross-cultural comparisons.
This study provides further evidence that Japanese tend to search for their weaknesses and shortcomings in an apparent effort to correct them. Self-enhancing motivations, routinely found within North American research, are elusive within a Japanese context. In contrast to motivations to find out what is good about themselves, Japanese appear more motivated to discover where they are not doing good enough. This cultural practice of self-criticism appears to serve Japanese in their quest to achieve connection and interpersonal harmony with others.
Baumeister, R. G., Tice, D. M., & Hutton, D. G. (1989). Self-presentational motivations and personality differences in self-esteem. Journal of Personality, 57, 547-577.
Bond, M. H., & Leung, K., & Wan, K. C. (1982). The social impact of self-effacing attributions: The Chinese case. Journal of Social Psychology, 118, 157-166.
Doi, T. (1986). The anatomy of self. Tokyo: Kodansha.
Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision of personal history. American Psychologist, 35, 603-618.
Heine, S. J., Kitayama, S., Lehman, D. R., Takata, T., & Ide, E. (1998). Self-enhancement and self-improvement: Two ways to motivate the self. Presentation at the Conference on Culture and Psychology, Stanford University. Aug. 19, 1998.
Heine, S. J., & Lehman, D. R. (1995a). Cultural variation in unrealistic optimism: Does the West feel more invulnerable than the East? Journal of Personality and Social Psychology, 68, 595-607.
Heine, S. J., & Lehman, D. R. (1995b). Social desirability among Canadian and Japanese students. Journal of Social Psychology, 135, 777-779.
Heine, S. J., & Lehman, D. R. (1997a). The cultural construction of self-enhancement: An examination of group-serving biases. Journal of Personality and Social Psychology, 72, 1268-1283.
Heine, S. J., & Lehman, D. R. (1997b). Culture, dissonance, and self-affirmation. Personality and Social Psychology Bulletin, 23, 389-400.
Heine, S. J., & Lehman, D. R. (in press). Culture, self-discrepancies, and self-satisfaction. Personality and Social Psychology Bulletin.
Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1998). Is there a universal need for positive self-regard? Manuscript submitted for publication.
Kashima, Y., Siegal, M., Tanaka, K., Kashima, E. S. (1992). Do people believe behaviors are consistent with attitudes? Towards a cultural psychology of attribution processes. British Journal of Social Psychology, 31, 111-124.
Kitayama, S. (1998). Self-criticism in Japan: Critical self-appraisal or modest self-presentation? Unpublished manuscript. Kyoto University.
Kitayama, S., Markus, H. R., Matsumoto, H., & Norasakkunkit, V. (1997). Individual and collective processes in the construction of the self: Self-enhancement in the United States and self-criticism in Japan. Journal of Personality and Social Psychology, 72, 1245-1267.
Kitayama, S., Masuda, T., & Lehman, D. R. (1998). Cultural psychology of social inference: The correspondence bias largely vanishes in Japan. Manuscript in preparation. Kyoto University.
Kitayama, S., Takagi, H., & Matsumoto, H. (1995). Causal attribution of success and failure: Cultural psychology of the Japanese self. Japanese Psychological Review, 38, 247-280 (in Japanese).
Lebra, T. S. (1976). Japanese patterns of behavior. Honolulu: University of Hawaii Press.
Markus, H. R., & Kitayama, S. (1991a). Cultural variation in the self-concept. In G. R. Goethals & J. Strauss (Eds.), Multidisciplinary perspectives on the self (pp. 18-48). New York: Springer-Verlag.
Markus, H. R., & Kitayama, S. (1991b). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224-253.
Maslow, A. (1943). A theory of human motivation. Psychological Review, 50, 370-396.
Meijer, Z. Y., & Semin, G. R. (1998). When the self-serving bias does not serve the self: Attributions of success and failure in cultural perspective. Manuscript submitted for publication. University of Virginia.
Miller, J. G., (1984). Culture and the development of everyday social explanation. Journal of Personality and Social Psychology, 46, 961-978.
Morris, M. W., & Peng, K. (1994). Culture and cause: American and Chinese attributions for social and physical events. Journal of Personality and Social Psychology, 67, 949-971.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton: Princeton University Press.
Schwartz, J. M., & Smith, W. P. (1976). Social comparison and the inference of ability difference. Journal of Personality and Social Psychology, 34, 1268-1275.
Shedler, J., Mayman, M., & Manis, M. (1993). The illusion of mental health. American Psychologist, 48, 1117-1131.
Stevenson, H. W. (1995). The Asian advantage: The case of mathematics. In J. J. Shields, Jr. (Ed.), Japanese schooling: Patterns of socialization, equality, and political control. (pp. 85-95). University Park, PA: Pennsylvania State University Press.
Takata, T. (1987). Self-deprecative tendencies in self evaluation through social comparison. Japanese Journal of Experimental Social Psychology, 27 (1), 27-36 (in Japanese).
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193-210.
Tesser, A. (1988). Toward a self-evaluation maintenance model of social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 181-227). San Diego, CA: Academic Press.
Weisz, J. R., Rothbaum, F. M., & Blackburn, T. C. (1984). Standing out and standing in: The psychology of control in America and Japan. American Psychologist, 39, 955-969.
Preparation of this manuscript was supported by fellowships from the Social Sciences and Humanities Research Council of Canada and the Japanese Society for the Promotion of Science to Heine, and by grants from the Social Sciences and Humanities Research Council of Canada to Lehman.
The study reported here was included in Heine’s dissertation, which was supervised by Lehman. We thank the other members of the Ph.D. committee, Jennifer Campbell, John Pinel, and Jim Russell, for their valuable feedback. We are especially grateful to Maho Harada, Shinobu Kitayama, and Nariko Takayanagi for their help in conducting this study.
Correspondence regarding this article should be addressed to Steven J. Heine at the Department of Psychology, 3815 Walnut Street, University of Pennsylvania, Philadelphia, PA 19104. E-mail may be sent to firstname.lastname@example.org.
Table 1. Means and Standard Deviations for Dependent Variables.
“Below Average” Decision
“Above Average” Decision
Accuracy of ICC Test
Note: Standard deviations are reported in parentheses.
Rows with different subscripts are significantly different at p < .05.
* Percentages are significantly different from 50% at p < .005.