Culture, Gender, and Self: A Perspective From Individualism-Collectivism Research

Abstract

Individualism and collectivism are often equated with independent vs. interdependent, agentic vs. communal, and separate vs. relational self-construals. Although these same concepts have been used to characterize both cultural and gender differences, a perspective of cultural evolution suggests it is unlikely. A division of labor within society may produce gender differences, but this cannot explain cultural differences. A study of self-construal involving 5 cultures (Australia, the United States, Hawaii, Japan, and Korea) shows that differences between these cultures are captured mostly by the extent to which people see themselves as acting as independent agents, whereas gender differences are best summarized by the extent to which people regard themselves as emotionally related to others. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Full text

Culture, Gender, and Self:
Is Women’s Sphere Universally Familial, and
Men’s Sphere Universally Societal?
Yoshihisa Kashima1, Emiko Kashima2,
Chi-Yue Chiu3, Tom Farsides4, Michele Gelfand5, Ying-Yi Hong6, Uichol Kim7,
Fritz Strack8, Lioba Worth8, Masaki Yuki9, and Vincent Yzerbyt10
1 Department of Psychology, The University of Melbourne
2 Discipline of Psychology, Swinburne University of Technology
3 Department of Psychology, The University of Hong Kong
4 School of Social Science, Sussex University
5 Department of Psychology, University of Maryland
6 Division of Social Science, Hong Kong University of Science and Technology
7 Department of Psychology, Chung-Ang University
8 Department of Psychology, University of Wuerzburg
9 Department of Behavioral Science, Hokkaido University
10 Department of Psychology, Catholic University of Luvain
Submitted to the Journal of Personality and Social Psychology
Running Head: Culture, Gender and Context-Sensitive Self
Correspondence:
Yoshihisa Kashima
Department of Psychology
The University of Melbourne
Parkville, Vic 3010
Australia
Phone: +613-8344-6312
Fax: +613-9347-6618
Email: ykashima@unimelb.edu.au
1

Abstract
Cross and Madson (1997) and Baumeister and Sommer (1997) suggested that, in North
America, women’s sphere may be familial, and men’s, societal. Yet, in East Asia, this may
not be the case because men’s life may be more connected to their families than women’s. It
is men’s, rather than women’s, duty to ensure that respects be paid to their ancestors and
prosperities of their offspring be maintained. The results of a cross-cultural study show that
men’s sphere is societal and women’s familial in five Western European cultures (Australia,
Belgium, Germany, UK, and USA). However, East Asian (Hong Kong, Japan, and Korea)
men’s self is more strongly connected to their families than their female counterparts, but it is
women who are more likely than men to subjugate their personal goals to the societal ones.
Although the context-sensitivity of self is one of the universal facts of life, the pattern of
context-sensitivity in gendered self depends on culture.
2

Culture, Gender, and Self:
Is Women’s Sphere Universally Familial, and Men’s Sphere Universally Societal?
Yoshihisa Kashima, Emiko Kashima, Chi-Yue Chiu, Tom Farsides, Michele Gelfand, Ying-
Yi Hong, Uichol Kim, Fritz Strack, Lioba Worth, Masaki Yuki, and Vincent Yzerbtyt
A number of theorists (e.g., Bakan, 1966; Gilligan, 1982) postulated a global gender
difference in self-conception, suggesting that men’s and women’s self-conceptions are
fundamentally different from each other. However, more recent theorists have emphasized the
context-sensitive nature of the relationship between gender and self. Deaux and Major (1987),
in a seminal paper on gender and social behavior, argued that the activation of men’s and
women’s gender-relevant self-conceptions depends fundamentally on the others with whom
they are interacting in the context. Depending on those others’ gender belief systems and
conceptions of the actors, as well as relevant situational cues, the actors may or may not act
on their gender. Gendered self is fundamentally context sensitive (see Smith, Noll, & Bryant,
1999, for an empirical demonstration in North America). Although there is some evidence to
suggest that gender differences in self vary systematically across contexts in North America
(see below), little is known about the role of culture in the shaping of gendered self in context.
As we point out later, there is theoretical reason to postulate the presence of a cultural
difference in gendered self in context. This paper examines this issue in a cross-cultural study
involving three English speaking (Australia, UK, & USA), two continental European
(Belgium & Germany), and three East Asian (Hong Kong, Japan, & Korea) cultures.
3

Culture, Gender, and Context-Sensitive Self
In North America, Cross and Madson (1997) and Baumeister and Sommer (1997)
provided specific hypotheses about gender differences in context-sensitive self-conceptions,
that is, how men’s and women’s self-conceptions are different in what way in what contexts.
These theorists’ writings jointly suggest that women’s sphere is familial, but men’s is societal.
In particular, Cross and Madson (1997) suggested that there may be a difference between
men’s and women’s self-conceptions along the dimensions of independent and interdependent
self-construal (Markus & Kitayama, 1991). Although both men’s and women’s self-construals
emerge out of gender socialization experiences, according to Cross and Madson, men’s self-
construals emphasize their separateness from others, while women’s self-construals stress
their connectedness with others, incorporating representations of others into their own (also
see Cross, Bacon & Morris, 2000). Men’s greater tendency to engage in aggression and
expression of power relative to women, for instance, is interpreted as indicating men’s
separateness from others.
In contrast, Baumeister and Sommer (1997) argued that there is a universal need for
social belonging, and that this should not differ between men and women. Instead, they
suggested, women may express their sense of belonging and connectedness within intimate
social contexts such as family and close relationships, whereas men do so within larger
societal contexts of community and society at large. In this view, even men’s greater tendency
for aggression and expression of power may be interpreted as men’s way of expressing their
connectedness with larger society. Consistent with these general theorizing of gender
differences, Gabriel and Gardner (1999) reported that in the US, men’s self-conceptions
tended to show greater connectedness with large groups and collectives such as country,
4

fraternity, and the like than women’s, but women’s self-conceptions were more connected
with intimate others (e.g., someone’s best friend, someone’s daughter).
Although this pattern of gender differences in context may be found in North America,
it remains to be seen whether the same pattern is found in other cultures. Although the past
empirical research focused on global differences in men’s and women’s self-conceptions
across cultures without paying close attention to social contexts in which gender differences
are expressed (e.g., Kashima et al., 1995; Li, 2002; Watkins et al., 1998a, 1998b; Watkins,
Mortazavi, & Trofimova, 2000; Williams & Best, 1990), there is a theoretical reason to
suspect that the same pattern of gender differences in context may not occur universally. As a
number of theorists suggested in the past (e.g., Geertz, 1984; Smith, 1978/1991; Taylor,
1989), self-conceptions are constructed on the basis of the cultural conceptions of the person
available in a given culture and historical period. In relation to gender, what it means to
become a man or a woman, or culturally presupposed trajectories of men’s and women’s life
courses, could vary significantly across cultures (e.g., Best & Williams, 1997; Bussey &
Bandura, 1999). In particular, different societies have different cultural models about how
boys and girls are born into their families and then grow up to become adult men and women
in larger society. Against the backdrop of different cultural models of gender socialization,
men’s and women’s self-conceptions may differ across cultures.
In particular, there may be some differences in how men and women are expected to
relate to family and society between East Asian and Western European cultures. In Western
European cultures, and especially in the United States, men are traditionally expected to leave
their parental homes (families) and establish themselves as independent contributors to
society. In this sense, Western men are expected to separate themselves from their families,
and to connect with society at large. However, Western women have traditionally been
5

expected to stay within and to connect with family. To put it differently, as the previous
literature in North America has shown, women’s self-conceptions may be more socially
connected than men’s in intimate social contexts, but men’s self-conceptions may indeed be
more socially connected than women’s in large scale collective contexts.
In contrast, East Asian models of gender socialization suggest a pattern somewhat
different from those prevalent in Western European cultures. In East Asia, men are to grow up
to be responsible for their family and lineage, and their role in society may be largely played
out through their family lineage as a “man of the family.” It is men’s, especially the eldest
son’s, responsibility and obligation to provide financially and socially for their aging parents
and growing families, and ensure the success and prosperity of their future generations.
However, East Asian women have traditionally been expected to leave their parental homes
and enter into their husbands’ families (e.g., Hsu, 1981; Lebra, 1976). In this way, East Asian
men may express their social connectedness with family more than East Asian women within
the family context.
Present Study
Men’s and women’s context-sensitive self-conceptions may differ across cultures, and
in particular between Western European and East Asian cultures. To test these contentions,
men’s and women’s self-conceptions in family, friendship group, and societal contexts were
examined in eight cultures: three English speaking (Australia, United Kingdom, and USA),
two continental European (Francophone Belgium, Germany), and three East Asian cultures
(Hong Kong, Japan, and Korea).
To cover the domains of self-conceptions that are sensitive to cultural and gender
differences, we used modifications of the measures adopted by Kashima et al. (1995) in their
6

examination of the nexus of culture, gender and self. The measure tapped a context specific
self-concept which examined whether the self’s or a particular ingroup’s goal is emphasized
(Yamaguchi, 1994). The ingroups were family, friendship group, or society at large in the
present study. Kashima et al. (1995) showed that this scale consisted of three subscales: two
individualist self scales (agency and assertiveness) in which the self was seen to do (agentic)
and say (assertive) things independently of his or her ingroup, and one collective self scale
that measured the extent to which one’s personal goal was regarded as secondary to one’s
ingroup’s goal, emphasizing obligations to the ingroups instead.
We expected that the extent to which one’s connectedness with a social sphere was
indexed by context-specific collective self, which taps the self-ingroup connectedness.
Although this conceptualization is somewhat different from Gabriel and Gardner’s (1999),
this is an alternative way to conceptualize and operationalize context-dependent social
connectedness. If we obtain a pattern similar to their findings within our framework, it should
provide a stronger conceptual replication and triangulation of Cross and Madson’s (1997) and
Baumeister and Sommer’s (1997) theorizing. It was hypothesized that the pattern of gender x
context interaction may differ across cultures, and therefore culture should moderate gender x
context interaction, yielding a three-way culture x gender x context interaction effect on
collective self. More specifically, in Western cultures, women may be more collective than
men in the family context, but men may be more collective than women in the societal
context. However, this pattern may not hold in East Asia. East Asian men may be more
collective than their female counterparts in the family context. Gender differences in the
societal context were not explicitly hypothesized.
Second, we expected that in line with the past research on individualism and
collectivism (Kashima et al., 1995; Markus & Kitayama, 1991; Oyserman, Coon, &
7

Kemmelmeier, 2002; Rhee, Uleman, Lee, & Roman, 1995; Triandis, 1995), Western selves
would be more individualist than East Asian ones. Oyserman et al.’s meta analysis found a
clear cultural difference between the USA and some East Asian cultures (Japan and Korea) on
individualism, but not on collectivism. However, there may be a cultural difference in context
effects on individual self as suggested by the past research (e.g., Rhee, Uleman, & Lee, 1996;
Triandis, McCusker, & Hui, 1990). No clear gender effects were hypothesized on individual
self.
Method
Participants
Three East Asian cultures (Japan, Korea, and Hong Kong), three English speaking
cultures (Australia, UK, and USA), and two continental European cultures (Belgium and
Germany) were sampled. Participants were all unmarried undergraduate students: 85 (41 men
& 44 women; 19.8 years old) from Seoul, Korea, 105 (53 men & 52 women; 20.0 years old)
from Tokyo, Japan, 77 (27 men & 50 women; 20.1 years old) from Hong Kong, 77 (27 men
& 50 women; 18.8 years old) from Melbourne, Australia, 80 (26 men & 54 women; 20.4
years old) from Maryland, the USA, 70 (12 men & 58 women; 20.3 years old) from Brighton,
the UK, (35 men & 40 women; 21.4 years old) from Würzburg, Germany, and 87 (44 men &
43 women; 20.8 years old) from Louvain la Neuve, Belgium (Francophone). Participants
were randomly assigned to three conditions with different contexts: family, friendship group,
and society.
Measures
The measures were administered as part of a larger study. Demographic variables
including age, gender, and ethnicity were examined. Some participants in Australia, the
8

United States, and the United Kingdom identified their ethnicity as not Western European
(e.g., Asian). These participants were excluded from further analyses. In addition, one
question was asked to gauge the participants’ subjective estimate of their socio-economic
status in their country. They were asked to state whether the income of the main income
earner of their family is highly above (1), above (2), about (3), below (4), or highly below (5)
the national average. All analyses were conducted with age and self-reported income as
covariates. As there was no effect of self-reported income on the results, but age correlated
with some of the measures, the results with only age as a covariate will be reported.
Nonetheless, even when age was not included as a covariate, the results remained the same.
To examine self-conceptions in the family, peer group, and societal contexts, 21 of the
items that Kashima et al. (1995) used to measure cultural differences in self-conception were
included with some modification. In the original study, the items generally described a
situation in which goals of a “group” and one’s own goals were in conflict with each other. In
the present study, the phrase “group” was defined in three different ways as family, group of
friends, or society in the relevant conditions. Of the 21 items, fourteen pertained to the
individual self, with seven tapping agentic self and the other seven, assertive self. The
remaining seven items were included to measure collective self within the given social
context. Example items are as follows: for agency, “I act more on the basis of my own
judgment than on other people’s decision”; for assertiveness, “I assert my opposition when I
disagree strongly with other people”; and for collective self, “I am prepared to do things for
other people at any time, even when I have to sacrifice my own interest.” For more detailed
description of the items, see Kashima et al. (1995).
9

Results
Preliminary Analyses
The 21 item measure of self-concepts was subjected to a pan-cultural factor analysis.
As van de Vijver and Leung (1997) recommended, the items were first standardized within
each culture to avoid the confounding of correlations among items due to systematic cultural
mean differences. In addition, following the same reasoning, we standardized the items within
each condition separately for each culture. This meant that when conducting a factor analysis,
there were no mean differences among the conditions across cultures. The total of 21 items
were initially subjected to a principal component analysis: the eigenvalues declined from 3.3,
2.3, 1.7, to 1.2, 1.0, 1.0 and so on, clearly suggesting that there were three factors. Following
a varimax rotation, the factor loadings showed that, consistent with the previous research
(Kashima et al., 1995), the initial grouping of the items into the three aspects (agentic,
assertive, and collective) was generally adequate. A separate factor analysis within each
culture also yielded a factor structure generally congruent with the results of the pan-cultural
analysis. To gauge the internal coherence of each subscale, Cronbach’s α coefficient was
computed for each subscale using the original seven items within the overall sample and also
each culture. As reported in Table 1, the internal coherence was moderate (approximately .60
and above). The average scores of the seven items were used in subsequent analyses.
[Insert Table 1 about here]
Culture and Gender in Context
A multivariate analysis of variance (MANOVA) was conducted to examine the
moderating effect of culture on gender relationships in different contexts. The dependent
variables were agentic, assertive, and collective selves, and the independent variables were
culture (8 cultures), gender (male vs. female), and context (family, friends, and society).
10

Given the size of the sample, a conservative type I error rate (α = .01) was set for this analysis.
Main effects of culture and context were significant, F(7,1737.783) = 10.73, Wilks’s Λ = .71,
η2 = .11, and F(6,1210) = 4.45, Wilks’s Λ = .92, η2 = .04. In addition, a culture x context
interaction effect was also significant, F(42,1795.485) = 1.94, Wilks’s Λ = .88, η2 = .04. Most
importantly, and as expected, a culture x gender x context three-way interaction effect was
significant, F(42,1795.485) = 1.80, Wilks’s Λ = .89, η2 = .04. To examine the multivariate
effects more carefully, the results of univariate analyses were examined for each self-aspect
separately.
Collective self. A culture x gender x context interaction effect was significant only on
collective self, F(14,607) = 3.26, η2 = .07, p. < .001. This three-way effect is the only
interpretable effect though other lower-level effects were also significant: culture, context,
and culture x context effects, respectively, F(7,607) = 3.36, η2 = .04, F(2,607) = 11.25, η2 = .
04, and F(14,607) = 2.82, η2 = .06. The relevant means are reported in Table 2. It is
interesting to note that generally in Western European cultures, men tend to have a higher
level of collective self than women in the societal context, but this is reversed in family. In
contrast, in East Asian cultures, men tended to have a higher level of collective self than
women in the family context, but not so in society.
[Insert Table 2 about here]
One way to represent visually this qualitative cultural difference in gender x context
interaction was to use a multidimensional scaling technique (see Kashima et al., 1995). We
first constructed for each culture a vector whose elements are the mean levels of collective
self for men and women, then computed the Euclidean distance between every pair of culture
11

vectors, and conducted the ordinal multidimensional scaling on the distance matrix. The one
dimensional solution yielded an adequate fit (Stress = .16, R2 = .93). The three East Asian
cultures had the highest scale values (Korea = 1.27; Hong Kong = .66; Japan = .64) and the
other Western cultures had low scale values (UK = -2.07; Germany = -.98; Belgium = -.25;
USA = .33; and Australia = .39), clearly indicating the East-West contrast in the patterns of a
gender x context interaction effect.
To summarize the East-West cultural difference, an additional ANOVA was
conducted on collective self with the East-West cultural contrast as a factor together with
gender and target. The three-way interaction was significant, F(2,644) = 13.20, p < .001, η2
= .04. The relevant means are displayed in Figure 1. In Western Europe, women’s self-
conceptions were significantly more collective than men’s in the family context, t(126) =
-3.07, p < .01, but, this was reversed in East Asia, t(89) = 1.88, p < .05. In contrast, Western
men’s self-conceptions were more collective than their female counterparts in the societal
context, t(134) = 1.77, p < .05, but the opposite was the case in East Asia, t(86) = 2.88, p. < .
01. The t-test results were all one-tailed because all the hypotheses were directional.
[Insert Figure 1 about here]
Despite this systematic cultural difference in pattern of the gender x context
interaction, the size of this interaction effect appears to vary substantially across cultures. The
largest effect size was found in the UK (.21), followed by Germany (.17), Japan (.10), Hong
Kong and Korea (both .03), the USA (.01), and Australia and Belgium (both .00). To shed
light on this, an exploratory analysis was conducted to examine the relationship between
effect size for the gender x context interaction and Hofstede’s (1998) score of masculinity, a
12

cultural index of gender role differentiation, according to Hofstede. It is based on the extent to
which stereotypically masculine work values (e.g., career as opposed to family) were
endorsed in a country. There is a marginally significant relationship between masculinity and
effect size of the gender x context interaction. Spearman’s ρ was .62, p = .051, across the
eight cultures, implying that the effect size is greater when masculinity of a culture is higher.
Individual self: Agency and assertiveness. On both agency and assertiveness, culture
and context main effects were significant: F(7,607) = 15.78, η2 = .15, and F(7,607) = 23.28, η2
= .21, for culture, and F(2,607) = 6.18, η2 = .02, and F(2,607) = 6.28, η2 = .02, for context. A
culture x context interaction was significant only for agency, F(14,607) = 2.55, η2 = .06.
Clearly, agentic and assertive selves, that is, doing and saying things on their own, need to be
examined separately.
[Insert Table 3 about here]
The extent to which people regard themselves as agentic varied across contexts in
different ways in different cultures (Table 3). Generally, in many cultures, self-conceptions
were most agentic in the societal context. People thought they tended to do things on their
own when they are in a large societal context compared to when they are within intimacy
groups (family, friends, or both). In continental European cultures, however, this was not the
case, especially in Germany. Instead, they are most agentic in the context of friendship
groups. The opposite in pattern to these continental European cultures was Hong Kong and
Japan, which showed the lowest level of agency in the friendship context (also the UK and
Australia to an extent). By contrast, the culture x context interaction effect was not significant
for assertiveness, F < 1. Generally, people considered themselves to be most assertive when
13

they are with their family (M = 3.6), and least assertive in societal context (M = 3.4), with
their assertive self in context of friendship groups in between the two (M = 3.5).
Nonetheless, when agentic and assertive self was examined in each context across
cultures, a familiar pattern expected from research on individualism emerged, that is, a higher
level of individualism in Western European cultures relative to East Asian ones. To clarify
this cultural pattern, additional ANOVAs were conducted on agentic and assertive self
measures in which the East Asian cultures (Hong Kong, Japan, and Korea) were contrasted
with the other Western European cultures; also included were target and gender effects. The
relevant means, reported in Figure 2, showed that in all contexts East Asian cultures were
lower on individualist self-conceptions than the other Western cultures, F(1,644) = 56.90, η2 =
.08, for agency, and F(1,644) = 96.46, η2 = .13 for assertiveness.
[Insert Figure 2 about here]
Interestingly, these analyses clarified an intriguing difference between agentic and
assertive self. With regard to assertive self, people saw themselves as asserting their views
independently in the more intimate contexts (family and friendship) than in the societal
context, F(2,644) = 7.68, η2 = .02. However, the reverse was true about agentic self: people
regard themselves to be more agentic in the societal context than in the family or friendship
contexts, F(2,644) = 6.70, η2 = .02. In addition, there was a subtle East-West difference in
pattern for the context effect on agentic self. The increase in agency from the family and
friendship contexts to the societal context was gradual for Western European cultures, but the
increment from the intimate (family and friendship) to societal context was more marked in
East Asian cultures, F(2,644) = 2.90, p. = .056, η2 = .01.
14

Discussion
Both in the East and West, men’s and women’s self-conceptions change as their social
contexts change. The context-sensitivity of self is one of the fundamental insights of social
psychology. Since the classical writings of William James (1890/1950) and symbolic
interactionists (e.g., Mead, 1934), it has been argued and absorbed into the contemporary
theorizing of self and identity (e.g., Markus & Kunda, 1986; McGuire & McGuire, 1982;
Turner, 1987) that our conceptions of ourselves change as do our social contexts. Clearly,
gendered self is no exception. In all cultures we examined, there was a degree of context-
sensitivity in gender differences in self-conception though its size may vary cross-culturally.
However, the pattern of context-sensitivity in gendered self depends on culture. In Western
European cultures, women’s sphere may be familial, and men’s, societal. Yet, in East Asia,
this may not be so. The results show that East Asian men express their social connectivity
more strongly in the family context than their female counterparts, but it is women who are
more likely than men to subjugate their personal goals to the societal ones.
Collective Self in Context
The results generally conformed to our initial expectations. Following Cross and
Madson (1997) and Baumeister and Sommer (1997), we had anticipated the pattern of gender
differences in Western European cultures. We had also expected the gender difference in the
East Asian family context. We nonetheless did not foresee East Asian women’s higher level
of societal connectivity. Two speculations are offered here: one that draws on the general
social change that confronts female university students in East Asia, and the other that is
based on the traditional image of women in East Asia.
15

In all of the three cultures we examined, the decades of industrialization and economic
developments have made it more likely than before that female graduates from universities
would seek jobs and careers in society. The female university students anticipate their
departure from their parental homes, and entrance into the wider societal world. It may indeed
be the case that they feel the expectations placed on them as adult participants of their society,
and their sense of connectivity to the larger society may be more heightened than their male
counterparts, for whom the entrance into the society is an expected course of socialization.
Nevertheless, it is also interesting to point out that there are some cultural exemplars of
women who acted as an embodiment of the morality and ethics of society at large, thus
expressing the ideal of women’s societal obligations and strong connection with the larger
society in East Asia. For instance, in a Chinese popular story of the Three Kingdoms, a story
of conflict among ancient Chinese dynasties, the mother of a brilliant strategist, Xushu,
hanged herself to death after learning her son’s defection from a morally respectable king to a
disreputable king to care for his mother. It is said that she was ashamed of her son whose
moral judgment was clouded by his concern for his family. In a Japanese popular song,
Ofukurosan (Mother), a son praises his mother’s teaching that he should be useful for society.
Future research may need to pay closer attention to a historical variation in gendered self in
context, taking the contemporary societal change as well as more traditional cultural legacies
into consideration.
Individual Self in Context
Another set of new findings pertains to the dissociation of agentic and assertive selves.
As noted earlier, although they are both concerned with individualist conceptions of oneself
as placing the individual over the ingroup, agentic self does so in doing and in action, but
assertive self in saying and in discourse. The present study suggests that these two aspects
16

should be treated differently. People see themselves as increasingly more agentic, but less
assertive, as the context becomes less intimate. It is interesting to note that the context
dependency of assertivness is in line with Brown and Levinson’s (1987) politeness theory.
Clearly, Brown and Levinson’s claim that the relationship between social distance and
politeness (or a lack of assertiveness) is universal is borne out by the current data.
Although this type of general trend was found across all the cultures examined, there
were several important East-West differences. First of all, on both agentic and assertive
selves, Western participants were more individualist than Eastern counterparts regardless of
the context and gender. This is in line with the literature on individualism and collectivism
(e.g., Hofstede, 1980; Triandis, 1989, 1995; Oyserman et al., 2002) as well as independent
and interdependent self-construal (Markus & Kitayama, 1991). In this regard, however, in line
with Oyserman et al.’s (2002) meta-analysis, we found this cultural difference only with
individualist self, but not with collective self. As a number of researchers (e.g., Kashima,
2001; Oyserman et al., 2002; Triandis, 1995) noted, individualism and collectivism need to be
conceptualized separately in cultural psychology. Second, on agentic self (but not on assertive
self), we found a subtle cultural difference in context effect. In Western cultures, the change
in agency was a gradual increase from the family, and friendship, to societal context, but it
was a more marked step-like function in Eastern cultures. As suggested by Triandis,
McCusker, and Hui (1990), the boundary of one’s imtimate ingroup may be more clearly
demarcated in East Asia than in Western European cultures.
Methodological and Theoretical Issues
One methodological caveat may be in order here. It has always been argued that
cultural differences obtained in psychological measures may be due to a variety of
methodological artifacts including response set (van de Vijver & Leung, 1999; also see Ji,
17

Peng, and Nisbett, 2000). However, the methodological problem may not arise when one
looks for a cultural moderator effect as we did in the current study. Indeed, we sought and
found a cross-over interaction effect involving culture (i.e., culture x gender x context
interaction as hypothesized). The only case in which the methodological rival hypothesis
arises is with regard to individualist self where we found a culture main effect in which
Westerners were more individualist than East Asians. Although this is in theory open to the
methodological rival hypothesis, it is unlikely to pose a serious threat. In the end, it is the
question of construct validity of measurement procedures, and the strongest construct
validation comes from the conformity of measures to theoretical expectations. In the present
study, the East-West cultural differences conformed to our theoretical expectations about
individualism. Of course, had the results contradicted the individualism-collectivism theory
and literature, the methodological rival hypothesis could have posed a major threat to our
inquiry. However, it was not the case.
Finally, the present research again underscored the importance of keeping the
conceptual separation between culture and gender effects. On the central psychological
dimensions in which we found cultural differences (i.e., individualist self), we did not find
gender differences, suggesting clearly that national cultures have psychological effects on self
that are distinct from gender. As Kashima et al. (1995) noted, men are not like Westerners, or
women like Easterners. Rather, there appears to be a gender-based division of labor in each
culture despite a variation in the extent to which such a differentiation exists. Note that the
spheres of social connectivity were divided between men and women (familial vs. societal) in
both East and West, although the pattern of this division of social labor was almost the mirror
image of one another. Viewed this way, it may be productive to conceptualize gender
18

differences as a kind of within-societal division of social labor and role differentiation (Eagly,
1987), which culture affords men and women in every society.
19

References
Bakan, D. (1966). The duality of human existence: Isolation and communion in
Western man. Boston, MA: Beacon Press.
Baumeister, R. F., & Somner, K. L. (1997). What do men want? Gender differences
and two spheres of belongingness: Comment on Cross and Madson (1997). Psychological
Bulletin, 122, 38-44.
Best, D. L. & Williams, J. E. (1997). Sex, gender, and culture. In J. W. Berry, M. H.
Segall, & C. Kagicibasi (Eds.), Handbook of cross-cultural psychology (2 nd
ed.) Vol. 3 (pp.
163-212). Boston, MA: Allyn and Bacon.
Brewer, M. B. & Gardner, W. (1996). Who is this “we”? Levels of collective identity
and self representations. Journal of Personality and Social Psychology, 71, 83-93.
Brown, P. & Levinson, S. C. (1987). Politeness: some universals in language use.
Cambridge, UK: Cambridge University Press.
Bussey, K. & Bandura, A. (1999). Social cognitive theory of gender development and
differentiation. Psychological Review, 106, 676-713.
Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender.
Psychological Bulletin, 122, 5-37.
Cross, S. E., Bacon, P. L., & Morris, M. L. (2000). The relational-interdependent self-
construal and relationships. Journal of Personality and Social Psychology, 78, 191-198.
Deaux, K. & Major, B. (1987). Putting gender into context: An interactive model of
gender-related behavior. Psychological Review, 94, 369-389.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation.
Hillsdale, NJ: Erlbaum.
20

Gabriel, S. & Gardner, W. L. (1999). Are there “his” and “hers” types of
interdependence? The implications of gender differences in collective versus relational
interdependence for affect, behavior, and cognition. Journal of Personality and Social
Psychology, 77, 642-655.
Geertz, C. (1984). “From the native’s point of view”: On the nature of anthropological
understanding. In R. A. Shweder & R. A. LeVine (1984). Culture theory (pp. 123-136).
Cambridge, UK: Cambridge University Press.
Gilligan, C. (1982). In a different voice: Psychological theory and women’s
development. Cambridge, MA: Harvard University Press.
Hamaguchi, E. (1985). A contextual model of the Japanese: Toward a methodological
innovation in Japan studies. Journal of Japanese Studies, 11, 289-321.
Hamaguchi, E. (1987). Nihonjin no kihonteki kachikan ni kansuru jikken-chosa
kenkyu [Experimental and survey research on the Japanese basic value]. Unpublished
manuscript, Osaka University.
Hermans, H. J. M. (1996). Voicing the self: From informaiton processing to dialogical
interchange. Psychological Bulletin, 119, 31-50.
Hofstede, G. (1980). Culture’s consequences. Thousand Oaks, CA: Sage.
Hofstede, G. (Ed.) (1998). Masculinity and femininity: The taboo dimension of
national cultures. Thousand Oaks, CA: Sage.
Hsu, F. L. K. (1981). Americans and Chinese. Honolulu, HI: University of Hawaii
Press.
James, W. (1890/1950). The principle of psychology. New York: Dover. [Original
work published in 1890].
21

Ji, L. J., Schwartz, L. J., & Nisbett, R. E. (2000). Culture, autobiographical memory,
and behavioral frequency reports: Measurement issues in cross-cultural studies. Personality
and Social Psychology Bulletin, 26, 585-593.
Kashima, Y. (2001). Culture and social cognition: Towards a social psychology of
cultural dynamics. In D. Matsumoto (Ed.), Handbook of culture and psychology (pp. 325-
360). New York: Oxford University Press.
Kashima, Y., Yamaguchi, S., Kim, U., Choi, S.-C., Gelfand, M. J., & Yuki, M. (1995).
Culture, gender, and self: A perspective from individualism-collectivism research. Journal of
Personality and Social Psychology, 69, 925-937.
Lebra, T. S. (1976). Japanese patterns of behavior. Honolulu, HI: University of Hawaii
Press.
Li, H. Z. (2002). Culture, gender and self-close-other(s) connectedness in Canadian
and Chinese samples. European Journal of Social Psychology, 32, 93-104.
Markus, H. R. & Kitayama, S. (1991). Culture and the self: Implications for cognition,
emotion, and motivation. Psychological Review, 98, 224-253.
Markus, H. R. & Kunda, Z. (1986). Stability and malleability of the self-concept.
Journal of Personality and Social Psychology, 51, 858-866.
McGuire, W. J., & McGuire, C. V. (1982). Significant others in self space: Sex
differences and developmental trends in social self. In J. Suls (Ed.), Psychological erspectives
on the self (Vol. 1, pp. 71-96). Hillsdale, NJ: Erlbaum.
Mead, G. H. (1934). Mind, self, and society. Chicago: University of Chicago Press.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). How American is
individualism? Relational Americans and other lessons from cultural and cross-cultural
research. Psychological Bulletin, 128, 3-72.
22

Rhee, E., Uleman, J., & Lee, H. K. (1996). Variations in collectivism and
individualism by ingroup and culture: Confirmatory factor analysis. Journal of Personality
and Social Psychology, 71, 1037-1054.
Rhee, E., Uleman, J. S., Lee, H. K., & Roman, R. J. (1995). Spontaneous self-
descriptions and ethnic identities in individualistic and collectivistic cultures. Journal of
Personality and Social Psychology, 69, 142-152.
Smith, C. J., Noll, J. A., & Bryant, J. B. (1999). The effect of social context on gender
self-concept. Sex Roles, 40, 499-512.
Smith. M. B. (1978/1991). Perspectives on selfhood. In M. B. Smith (Ed.), Values,
self & society: Toward a humanist social psychology. New Brunswick, NJ: Transaction
Publishers. [Original published in American Psychologist, 33, 1053-1063.
Taylor, C. (1989). Sources of the self. Cambridge, MA: Harvard University Press.
Triandis, H. C. (1989). The self and social behavior in differing cultural contexts.
Psychological Review, 96, 506-520.
Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview.
Triandis, H. C., McCusker, C., & Hui, C. H. (1990). Multimethod probs of
individualism and collectivism. Journal of Personality and Social Psychology, 59, 1006-1020.
Turner, J. C. (1987). Rediscovering the social group. Oxford, UK: Blackwell.
van de Vijver, F. & Leung, K. (1997). Methods and data analysis for cross-cultural
research. Thousand Oaks, CA: Sage.
Watkins, D., Mortazavi, S., & Trofimova, I. (2000). Independent and interdependent
conceptions of self: An investigation of age, gender, and culture differences in importance and
satisfaction ratings. Cross-Cultural Research, 34, 113-134.
23

Watkins, D., Adair, J., Akande, A., Cheng, C., Fleming, J., Gerong, A., Ismail, M.,
McInerney, D., Lefner, K., Mpofu, E., Regmi, M., Singh-Sengupta, S., Watson, S., Wondimu,
H., & Yu, J. (1998a). Cultural dimensions, gender, and the nature of self-concept: A fourteen-
country study. International Journal of Psychology, 33, 17-31.
Watkins, D., Adair, J., Akande, A., Gerong, A., McInerney, D., Sunar, D., Watson, S.,
Wen, Q. F., & Wondimu, H. (1998b). Individualism-collectivism, gender and the self-
concept: A nine culture investigation. Psychologia, 41, 259-271.
Williams, J. E. & Best, D. L. (1990). Sex and psyche: Gender and self viewed cross-
culturally. Newbury Park, CA: Sage.
Yamaguchi, S. (1994). Collectivism among the Japanese: A perspective from the self.
In U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon (Eds.), Individualism and
collectivism (pp. 175-188). Newbury Park, CA: Sage.
24

Author Note
The empirical study reported in this paper was supported by a grant from the Australian
Research Council to the first two authors: the first author, Y. Kashima, and the second author,
E. S. Kashima. The authorship of the other authors was equally shared; the order of their
appearance is alphabetical, and does not reflect differences in the amount of contribution. We
thank Debbie Best for her useful comments on an earlier draft of the paper. Correspondence
should be directed to Yoshi Kashima, Department of Psychology, The University of
Melbourne, Parkville, Victoria 3010, Australia (Email: ykashima@unimelb.edu.au).
25

Table 1. Cronbach’s alpha coefficients for each subscale (agentic, assertive, and collective) of the self-
measures for the overall sample and each culture.
All HK Japan Korea Aust. UK USA Belg. Germany
Agentic .73 .84 .72 .72 .54 .64 .73 .56 .67
Assertive .71 .65 .64 .53 .66 .62 .70 .60 .77
Collective .62 .66 .58 .78 .57 .69 .64 .46 .63
26

Table 2. Mean levels of men’s and women’s collective self in family, friend, and societal contexts in
Hong Kong, Japan, Korea, Australia, UK, USA, Belgium and Germany.
Context Gender HK Japan Korea Aust. UK USA Belg. Germany
Family Men 3.52 3.55 3.89 3.43 2.21 3.57 3.14 2.85
Women 3.32 3.17 3.82 3.38 3.67 3.74 3.23 3.40
Friends Men 3.68 3.14 3.55 3.71 3.26 3.33 3.51 3.37
Women 3.76 3.58 3.27 3.59 3.54 3.17 3.48 2.92
Society Men 3.03 2.88 3.18 3.26 3.29 3.23 2.97 3.49
Women 3.27 3.32 3.48 3.16 2.88 3.14 3.00 3.05
Note. The underlined means numerically conform to the expected pattern of gender differences. In
Western European cultures, men are more collective than women in society, but women are more
collective than men in family. In East Asian cultures, men are more collective than women in family.
No specific hypothesis was put forward about a gender difference for the friendship context, or for East
Asians in the societal context.
27

Table 3. Agentic and assertive self in family, friendship, and societal contexts in Hong Kong, Japan,
Korea, Australia, UK, USA, Belgium, and Germany.
Context HK Japan Korea Aust. UK USA Belg. Germany
Agentic Self
Family 2.65 3.39 2.50 3.35 3.33 3.02 3.18 3.35
Friends 2.38 3.03 2.70 3.31 3.25 3.49 3.41 3.55
Society 3.22 3.37 2.74 3.66 3.30 3.61 3.27 3.22
Assertive Self
Family 3.31 3.66 2.91 4.05 3.81 3.71 3.90 3.69
Friends 3.18 3.56 2.67 3.57 3.56 3.72 3.81 3.64
Society 3.09 3.20 2.72 3.59 3.55 3.56 3.83 3.65
28

Figure Caption
Figure 1. Mean levels of men’s and women’s collective self in family, friend, and societal
contexts in East Asian and Western European cultures.
Figure 2. Mean levels of agentic and assertive self in family, friend, and societal contexts in
East Asian and Western European cultures.
29

30
2.8
3
3.2
3.4
3.6
3.8
Family Friends Society
Men Women
Men
Women
2.8
3
3.2
3.4
3.6
3.8
Family Friends Society
Me n Wom en
Western EuropeanEast Asian

31
2.5
2.75
3
3.25
3.5
Family Friends Society
East West
3
3.25
3.5
3.75
4
Family Friends Society
East W est
East
West
Agentic Self Assertive Self