As a result of increasing immigration, American schools are now catering to a growing proportion of students who are immigrants and children of immigrants. Often these children are confronted with conflicting notions of exactly what constitutes their cultural values. Growing up in America, these cohorts of adolescents are exposed to a specific set of beliefs at school, but they may receive a contradictory set of values at home. These environmental, and cultural, influences are important in the development of the self. Compounded with the stress of the adolescent life stage, immigrant boys and girls face unique challenges that, perhaps, non-immigrant children do not.

Over the last four decades, adolescent self-esteem has been measured consistently by social psychologists. Recently in North America, the self-esteem of immigrants and of adolescents from other cultures has been measured and compared with American-born adolescents. The study of adolescent self-esteem is important because it directly relates to and is derived from the larger social structure wherein values are intrinsically transmitted. Classical and contemporary social psychologists agree that people have a pervasive need for self-esteem (Greenberg et al 913). Self-esteem theory and measurement inherently imply cultural and gendered assumptions. In other words, the implication is that self-esteem is an individual characteristic that all humans possess and are continually striving to improve. Improvement is conceptualized in terms of further self-development. Most importantly, at the very basis of all these assumptions, lies the crucial Western notion of the individuated and autonomous self. The majority of researchers believe that self-esteem lies on a linear vector: individuals who espouse high self-esteem are considered to be functioning smoothly in society whereas those with lower self-esteem are thought to be struggling.

This paper seeks to critically examine the theoretical assumptions about self-esteem and the scale commonly used to measure the concept of self-esteem in order to discern whether the assumptions and measurements can apply to adolescents from diverse cultural backgrounds where notions of self are distinctly different. I will distinguish between American-born children, immigrant children, and children who live in other countries.[1] Although I am writing generally about groups of adolescents, I recognize that diverse circumstances confront each gender and ethnic group. Adolescents do not form a homogenous group. I will argue that Western assumptions of the “self” are guiding social psychological research on self-esteem and are likely to be highly inappropriate for those whose identities manifest from non-Western cultures, or who blend or straddle multiple cultures. In our increasingly globalized and oft-times transnational world, the study of adolescent self-esteem must be critically reassessed for its cultural assumptions.

In this paper, I first give an historical overview of the concepts of self, self-esteem, and adolescence. Then, I critically analyze theoretical assumptions and methodological applications of self-esteem. Finally, I conclude that the existing assumptions of self-esteem reflect and are biased toward Westernized conceptions of the self that are not inclusive of diverse cultural norms that may present a less individuated and more collective notion of the self. I suggest that we refine our current universalistic notion of self-esteem to incorporate localized cultural diversity and gender socialization. Although the focus of this paper is the improvement of self-esteem theory and measurement through focusing on cultural diversity and gender socialization, additional markers of social difference, such as class and sexuality, deserve attention in future studies.

Historical Overview of the Study of Adolescence, Self, and Self-Esteem

When evaluating the life-stage of adolescence and the psychosocial components of self and self-esteem, it is vital to recognize that Western psychologists and sociologists socially constructed these concepts within the last century. Psychologist William James (1890) developed the concept of self-esteem, coincidentally, during the same historical period in which Stanley Hall (1904) coined the term adolescence. Adolescence was not perceived as a distinct stage of life in earlier centuries; individuals simply moved from childhood into young adulthood. The rise of the middle class through the period of the Industrial Revolution allowed many privileged teens to remain outside the labour force; thus, education became increasingly important (Henslin 73). As a result of these social changes, the term adolescence was created to indicate the gap between childhood and young adulthood (Henslin 73).

The study of self and self-esteem originated from a psychosocial perspective. The concept first arose in psychology and can be traced back to the writings of William James in the late 19th century. James was the first social scientist to develop a clear professional definition of the self (Turner 343). In his typology of self, James' description of the social self recognized that people's feelings about themselves arise from interaction with others; he recognized that humans have the capacity to view themselves as objects and to develop self-feelings and attitudes toward themselves (Turner 344). According to James,

Sociologists Charles Cooley (1902) and Herbert Mead (1934) expanded upon earlier studies in the psychosocial development of the self. These scholars and other symbolic interaction theorists emphasize the way the self is socially constructed in interaction, based on people's shared understandings of social roles, rules, symbols, and categories. Following this symbolic interactionist line of reasoning, the social construction of self, then, for immigrant adolescent girls and boys, is based on people's shared understandings of social roles, rules, symbols, and categories. But who are these people? In the lives of American-born adolescents, these people are often their parents/guardians, teachers, and peers. For American-born children, parents and teachers probably align quite closely with their traditional, mainstream interpretation of social roles, rules, symbols, and categories. Peers may adhere to different interpretations based on their own cultural upbringing. For adolescent immigrants, on the other hand, it is possible that all three categories (parents, teachers, and peers) are inconsistent with their interpretation of social roles, rules, symbols, and categories. Parents may adhere to a cultural interpretation from their original homeland that clashes with teachers who adhere to a mainstream American interpretation, while peers could posit alternative explanations. Thus, these earlier studies had a limited perspective and the notion of "shared understandings" needs to be questioned when taking into consideration the cultural influences encountered by all adolescents and, in particular, immigrants and children of immigrants.

Immigrants arrived in the United States (mainly from Mexico, the Philippines, Cuba, and El Salvador) in considerable numbers after the Immigration Act of 1965. This act ended the national origins quota system and opened the doors for individuals and their families to enter the United States as immigrants under various categories; it also allowed naturalized citizens to sponsor the immigration of their siblings and parents (Ramisetty-Mikler 36). In the same historical period as the passage of the Immigration Act, American social psychologists began to explore self-esteem in depth. Morris Rosenberg is a main contributor in the rebirth of self-esteem studies in social psychology, which had been dormant since the turn of the twentieth century (Mruk 13). Rosenberg's (1965) Society and the Adolescent Self-Image opened a new door to psychologists and sociologists and helped bridge a disciplinary gap with his self-esteem theory and his Rosenberg Self-Esteem Scale (See sidebar below). Rosenberg's self-esteem theory relies on two factors: (1) reflected appraisals and (2) social comparisons. Regarding reflected appraisals, Rosenberg acknowledges that

Rosenberg Self-Esteem Scale To what extent do you disagree or agree with the following statements? 1=Strongly Disagree 2=Somewhat Disagree 3=Slightly Disagree 4=Slightly Agree 5=Somewhat Agree 6=Strongly Agree On the whole, I am satisfied with myself. 1 2 3 4 5 6 At times I think I am no good at all. 1 2 3 4 5 6 I feel that I have a number of good qualities. 1 2 3 4 5 6 I am able to do things as well as most other people. 1 2 3 4 5 6 I feel I do not have much to be proud of. 1 2 3 4 5 6 I certainly feel useless at times. 1 2 3 4 5 6 I feel that I'm a person of worth, at least on a level equal with others. 1 2 3 4 5 6 I wish I could have more respect for myself. 1 2 3 4 5 6 All in all, I am inclined to feel that I'm a failure. 1 2 3 4 5 6 I take a positive attitude toward myself. 1 2 3 4 5 6 Source: Rosenberg, Morris. Society and the Adolescent Self-Image. Princeton: Princeton University Press, 1965.

In addition to reflected appraisals and social comparisons, a third construction is inherently subsumed in the theoretical literature: self-esteem as a basic human need. Embedded in this assumption is the connotation that high self-esteem equates to the healthy, happy individual while low self-esteem equates to the unsatisfied, disconcerted individual. I review these assumptions and conclude that they do not account for local cultural variation in the understanding of self. I contend, however, that identity theory does recognize the interplay of structural and cultural influences.

First, reflected appraisals signify that a person's self-esteem is a product of how that person believes others see her/him (Hughes and Demo 134). Second, the assumption of social comparisons is based on Festinger's (1954) social comparison theory that argues we evaluate ourselves, in part, by comparing ourselves with others. This suggests that persons belonging to low-status groups will internalize the negative evaluation of themselves by society and as a consequence have low self-esteem (Hughes and Demo 134).

However, more recent research indicates that these assumptions are inaccurate (Hughes and Demo 152; Crocker and Major 612). In theory, self-esteem should be lower in dissonant social contexts, that is, where the level of social dissimilarity is higher along with exposure to negative stereotypes and reflected appraisals about one's group of origin, whichever mechanisms of perpetual defense are deployed to protect self-esteem (Rumbaut 754). Rosenberg (xxv-xxvii) originally assumed that a subordinate group (i.e. African-Americans) would evaluate themselves against the dominant group (whites). North American researchers were perplexed after consistently finding that African-Americans demonstrate higher self-esteem than white Americans. Rather than seeing themselves through the eyes of those in the dominant white group, researchers found that African-American adolescents compare themselves relative to those in their own ethnic group. Members of stigmatized groups actually avoid threats to their self-esteem by comparing themselves primarily with others who are members of their own stigmatized group rather than with members of an advantaged out-group (Crocker and Major 143).

This finding suggests that American-born adolescents compare themselves with those in the same ethnic group. This is not possible for immigrants who do not have anyone else in their ethnic category with whom to compare themselves. My ethnographic research in a junior high school presents a relevant illustration (Flynn 2001). One afternoon when I was conducting research, I walked up to a girl sitting by herself. I asked her name and she said, "Sadhra." I then asked where she was from; she said, "India."[2] While doing fieldwork research for one year, I never saw Sadhra sitting with or interacting with any other girls or boys at lunch. She sat on the same concrete ledge in the far southeast portion of campus every day. There were no other Indian children at this school. Where does Sadhra's sense of self and self-esteem come from? Based on the findings of the reflected appraisals and social comparison assumptions, Sadhra does not have peers with whom to compare herself. The concept of identity, specifically that identities are arranged into hierarchies of salience, is useful. When asked, it became clear that Sadhra's identity is tied to her family, religion, and her friends in India. Neither the self-esteem assumptions nor a universal quantitative scale incorporate these dimensions of identity or cultural influence.

In addition to suppositions of reflected appraisal and social comparison, the assumption that self-esteem is a basic human need is problematic. In cultures that tend to be more individualistic, for example the United States, the discourse surrounding self-esteem (popular and academic) is that all individuals possess a self, and therefore, self-esteem. In nations with individuated cultural values, "the individual has a right and responsibility, in fact a moral obligation, to become separate, autonomous, efficacious, and in control" (Heine 760). According to literature directed at a popular audience, if one has low self-esteem, "we can find a way to increase it!"[3]

In contrast to individualist societies, collectivist cultures, such as Japan, honour the interdependent self. The interdependent self is most complete when the individual is seen as functioning smoothly within a larger, more encompassing collective (Heine 770). In Japan, the term 'self-esteem' did not even exist until recently. A self-critical orientation and an incomplete (ki ga sumanai) feeling in terms of performance are characteristics of Japanese culture (Heine 779). Heine explains:

Through this evaluation of self-esteem theory, it is clear that the construction of self-esteem as reflected appraisals, social comparisons, and a basic human need are biased culturally and influenced by a Western context. Rosenberg observes:

However, I argue that identity theory provides the missing link in the relationship between the self and the social structure. Where past studies defined the self as socially constructed through micro-interaction based on people's shared understandings, recent theorists are reconceptualizing the notion of self into more precise terms. Self is constructed from a number of identities that are arranged into hierarchies of salience and importance; the higher an identity is in the hierarchy, the more influential it is to an individual's presentation of self (Turner 374).

Sheldon Stryker is a social psychologist whose identity theory makes a valuable contribution the study of self-esteem. Stryker (1980) claims that the purpose of social psychology is to examine the link between micro-processes and macro-structure. His identity theory proposes that the concept of self is composed of a hierarchical set of identities, each of which is tied to roles within the social structure, thus making the link between self and society more explicit. An example of how identity theory and self-esteem are closely aligned is the work of Chodorow (1974). She argues that if individuals' social category is highly valued, they value themselves highly; whereas if their social category is of a lower status, they lose self-esteem.

Identity theory emphasizes ‘commitment’ as a means for conceptualizing the link between micro-interaction and macro-structure: when identity is intertwined with the norms and values of the broader society, self-esteem is even more dependent on their successful implementation (Turner 375). Thus, the way society is organized - individualist versus collectivist - influences cultural perceptions of the self, identity, and presumably self-esteem.

The following is an examination of commonly used self-esteem methods and a call for quantitative and qualitative methods that incorporate the principles of identity theory.

Self-Esteem Methodologies

From a quantitative standpoint, one of the most challenging issues in the study of self-esteem is the causal direction of the variable. Should self-esteem be an independent variable (a variable that causes certain behaviour(s) and/or attitude(s)) or a dependent variable (a variable that is influenced by, or a result of, behaviour(s) and/or attitude(s))? The answer to this question is based purely on the researcher's theoretical orientation. Thus, issues of validity and reliability are created for both sociologists and psychologists. Social psychologists who prioritize psychological components of self argue that self-esteem is an independent variable. In their theoretical view, self-esteem is an internal process. Social psychologists who emphasize sociological mechanisms of self, alternatively, contend that self-esteem is a dependent variable that is influenced by environmental and social processes. For the most part, social psychologists adhere to the notion that self is influenced by both internal and external processes but side with one or the other based on divisional affiliation. I align theoretically with the sociological perspective that self-esteem is influenced by environmental variables such as culture and discrimination.

As mentioned previously, Rosenberg's ten-item scale measures global self-esteem and remains the most-widely used scale of self-esteem by social psychologists today despite almost four decades having passed since its contrivance. Using this scale cross-culturally creates two methodological concerns: (1) comprehension and/or translation and (2) response bias. In the most comprehensive study of children of immigrants living in the United States to date, Portes and Rumbaut (2001) use the Rosenberg Self-Esteem Scale to measure the self-esteem of numerous youth. The results indicate that Mexicans and the Indochinese show the lowest self-esteem scores; the highest self-esteem scores were found among the respondents from Jamaica and the other English-speaking West Indies followed by Haiti, Cuba, and the other Latin American countries. The authors note that Rosenberg's scale could be partially invalid because of cultural peculiarities (224). However, in the description of methodology, Portes and Rumbaut do not mention issues involving comprehension. Indeed, all student questionnaires were in English, but issues concerning comprehension of the English language are not addressed other than an acknowledgement that limited bilingualism reduces self-esteem (Portes and Rumbaut 221). It seems possible that limited bilingualism does not necessarily reduce self-esteem; rather, it may lead to issues of comprehending the wording or meaning of questions. Statements, such as “I feel that I'm a person of worth, at least on a level equal with others,” have varied meanings depending on cultural norms. In some cultures, modesty may be a highly-valued trait, so one might disagree with the statement in order to appear humble.

Scott and Scott (1998) describe the difficulty of translation (from English into four other languages: Cantonese, Berman, Japanese, and Mandarin), in their cross-cultural study of adolescent adjustment:

Response bias findings relate to social value differences in collectivist (midpoint bias) and individualist (extreme point bias) cultures. An example is the assumption that self-esteem is a universal necessity. Heine et.al. point out that “such diametrically opposed evaluations of the value of self-confidence and self-esteem between cultures force us to question the universality of motivations to possess a positive self-view” (Heine et.al. 779). Heine et.al. also note that the cultural bias of the Rosenberg Self-Esteem Scale is evident in longitudinal studies of Canadian and Japanese students exposed to North American or Japanese culture. The longer persons of Asian descent spent participating in North American culture, the higher were their self-esteem scores; and the longer those of North American descent participated in Asian culture, the lower were their self-esteem scores ( 777). It appears that participating in a North American cultural context leads Japanese to attend more to their positive features as individuals and thus increases their overall positive self-evaluation; whereas, living in Japan seems to lead North Americans to experience a decrease in the positivity of their self-evaluations (Heine et.al. 777). This is also true for other ethnic groups. All immigrant groups in Portes and Rumbaut's (207) study increased their self-esteem over a three-year period. They explain the boost in self-esteem as a result of the increased length of acculturation toward American culture (De Las Fuentes and Vasquez 138).[5]

When studying self-esteem cross-culturally (including immigrants in the U.S.), it is necessary to be cognizant of translation issues and response bias and to address such methodological shortcomings. Quantitative studies lend themselves to generalizable data wherein the individual histories and cultural complexities of adolescents are not fully realized. To overcome quantitative shortcomings, qualitative research is imperative in order to gain a more nuanced understanding of the influence of culture in adolescent identity construction (see Kibria 1993 and Wolf 1997 for examples). The study of self-esteem is strengthened immeasurably by the use of a variety of research methodologies; whenever different research methodologies lead to similar findings we can be more confident about inferences (Covington 77).

It would be interesting to compare qualitative and quantitative studies that rely on identity theory versus studies that rely on traditional self-esteem theory. Identity theory presumes that individual salience hierarchies exist based on commitment to particular identities. Clearly, cultural values are part of such hierarchies and may have an impact on levels of self-esteem. Role identities are internalized role designations in which the societal and cultural expectations for a particular role, connected with a position in society, are internalized by the individual (Stryker and Burke 291). Commitment to the roles depends upon the quality of relationships formed by the individual through interaction in that role (Stryker and Burke 285). Perhaps such a study could include quantitative measures of how much the adolescent's sense of self worth is related to ethnic identity and attached to cultural values. Ethnic identity formation is more salient for children of immigrants than for non-immigrant children; often, for these children becoming an American includes the adoption or rejection of ethnic identity (Portes and Rumbaut 221). Prior to embarking on a quantitative study, the examination of various cultures and cultural values is necessary. Michael Burawoy (2000) advocates the use of the extended case method as a technique that incorporates both comparative-historical methods and ethnographic methods in global and cultural contexts. In the case of immigrants and adolescents of immigrants, integrating acculturation and how cultures are negotiated could also be part of the study.

In the following section, I propose (along with Carpenter and Johnson 2001, Knox 1998, and Smith 1999) that we borrow from findings regarding collectivist and individualist societal differences in self-esteem theory and measurement, and apply them to gendered self-esteem results.

Gendered Self-Esteem

After reading a refreshing cultural analysis of the self that compares North America and East Asia by Heine et.al., it was disconcerting that gendered distinctions were not a part of their analysis. These authors fell into the very trap that they criticize. They condemn North American scholars for imputing Western philosophy to East Asian countries, yet are they not doing the very same thing by imposing monolithic cultural attributes of the self both to girls and to boys? Self-esteem theory and measurement has not only a cultural bias but also may have a gender bias (Carpenter and Johnson 254, Knox 62, and Smith 281).

Cross-culturally and cross-nationally in the U.S., researchers (AAUW 1991; Rumbaut 1994; Scott and Scott 1998) find that girls have lower self-esteem than boys. In these studies, an interesting biologically-based explanation for the plummet of girls' self-esteem during adolescence is the onset of puberty. Findings indicate that puberty, at any age, generally has a positive socio-emotional and peer-relational effect for boys (Brooks-Gunn and Reither 39; Ge, Conger, and Elder Jr. 413). Puberty is viewed as a rite of passage and a time where a boy becomes a man. Weight gains are often associated with muscle gain, seen as a positive male identity marker.

Puberty for girls, on the other hand, has a much more complicated effect and this effect is shown most poignantly based on the onset age of puberty. Girls who mature at a younger age face far-reaching consequences. Studies suggest that they have more depressive symptoms and lower self-esteem (Ge et al 373). These consequences are not the same for on-time or late-maturing girls. Early-maturing girls are also more likely to report that they feel they are overweight, and experience more depression, than their on-time or late-maturing counterparts (Ge et al 373).

Given just this one example of gender-linked factors in adolescent self-esteem, a universal application of self-esteem methodology is decidedly questionable. Current definitions and methods of self-esteem that employ such universals may not adequately describe the adolescent female self-concept; therefore, the conclusions of past studies of gender differences in levels and expressions of self-esteem may be misleading (Knox 64).

In addition to factors such as puberty, individualist versus collectivist socialization along gender lines exist, linking the critiques in cross-cultural implications of self-esteem theory and those of gender. In this section, differences in cultural socialization between East Asian and North American countries are elaborated through a gender lens. Using the United States as a case study, I argue that boys are raised oriented toward individualism and girls are raised oriented toward collectivism, shedding additional light on the finding that girls consistently have lower self-esteem than boys.

In Japan, there is a shared belief in the interdependent self. "The self is not considered separate and autonomous; rather it is within the contextual fabric of individuals' social relationships, roles, and duties that the interdependent self most securely gains a sense of meaning" (Heine et.al. 770). Socialization in the Japanese education system revolves around hansei (self-reflection). Hansei involves the individual reflecting on particular events and focusing on what was not done ideally and what she or he should try to improve in the future (Heine et.al. 770). Heine et.al. (770-771) emphasize that children are culturally and socially encouraged to reflect upon their own inadequacies and that self-criticism is in service of future improvement and achievement of the self. Age is also a consideration in Japanese culture where a system of seniority exists, and achievement may be recognized only after years of practice. Japanese adolescents, in particular, because they are among the youngest (and thus lowest status) in the social system, are socialized to embody what North American social psychologists refer to as low self-esteem. This label of low self-esteem is a reflection of the masculine Western assumption that humans experience self-esteem as an imperative need in contrast to the cultural collectivist orientation that encourages self-criticism and the persistent striving for self-improvement in Japan.

In the United States, individualism is a prominent value where "children are socialized as agents through culturally significant images of stories of men who are masters of their fates and captains of their ships, lone cowboys, and boys who pull themselves up by their own bootstraps" (emphasis added, Heine et.al. 769). What is missing in this quote is that all children are not socialized in the same way. American boys are socialized to be independent while American girls are socialized to be interdependent in a way that corresponds to the difference between North American and East Asian cultures. Following is a brief analysis of ways in which American children are socialized to adhere to disparate cultural values.

Gender is culturally and socially constructed. Socialization is the lifelong process by which a society's constructed values and norms, including those pertaining to gender, are taught and learned (Renzetti and Curran 58). Gender schema theory (Bem 1981) argues that as children learn appropriate cultural definitions of gender, this becomes the key structure around which all other information is organized. Therefore, much of the socialization process for children involves learning how to act like a "boy" and learning how to act like a "girl," and in what ways these genders differ. Gender polarization is exaggerated through stereotyping and reinforced by the media, clothing, books, toys and games. American stereotypes of masculinity include independence, strength, and dominance. American stereotypes of femininity include interdependence, sensitivity, passivity, and emotionality. Children are commonly socialized along gender stereotypical lines. Often, the first question after a baby's birth from eager parents is, "Is it a boy or a girl?" The answer, in most cases, determines how that child is treated and is the commencement of gender construction (Lorber 14; Renzetti and Curran 69).

Additional examples of how parents influence gender socialization are the books, toys, and media that are chosen for children. For instance, studies (Clark 1993; Flynn 2003; Gooden and Gooden 2001) find that children's picture books play an important role in gender socialization by providing an indication of social norms. Researchers emphasize that gendered messages in children's media encourage boys to be practical while girls are encouraged to be emotional (Renzetti and Curran 75). Such findings do not support the notion that children are raised to be individualistic in the United States, rather, these results suggest that boys are raised as individualistic. Toys, cartoons, and games enhance this gender division. Parents treat boys and girls dissimilarly through the encouragement of gender-typed behavior and communication (Renzetti and Curran 71-72). Girls tend to be given dolls while boys are given action figures. Some of the lessons that boys learn from toys are to be strong, adventurous, active, and heroic; girls learn to be domestic, play cooperatively, and how to beautify themselves.

It is important to note that gendered roles are in the process of changing. Fathers are more involved in childrearing practices and domestic activities, children are wearing unisex clothing, and men and women are reaching similar levels of education and working at the same jobs (Lorber 14). However, equality has not been achieved. Gender inequality is maintained through practices of boundary maintenance that uphold differences between dominant and subordinate groups (Lorber 32; Schwalbe 430-431; Thorne 86).

Global self-esteem measures mask important differences within the domains of self-concept that contribute to self-esteem (Knox 61). As indicated in the Japanese case study, culture influences self and systematically biases responses to Likert measurement scales. Is this response bias (i.e. measurement error) or true substantive differences that are incorrectly interpreted? Could differences in self-esteem be the result of a gendered response bias? Following the above outlined logic, cultural socialization findings in America indicate that girls are socialized toward a collectivist orientation while boys are socialized toward an individualist orientation. This may account for a gendered response bias induced methodologically. It is possible that girls may be more likely to use the midpoint because of their socialization, like East Asian cultures, and boys may be more likely to use the extreme points, similar to North American cultures. It may be probable that girls and adolescents from other cultures are more likely to choose the midpoint (slightly agree, slightly disagree) because they do not want to contravene appropriate gender roles - such as appearing "stuck up" or acting as if they are better than others.

Previous research (i.e. Carpenter and Johnson 2001, Knox 1998, and Smith 1999) using the collective self-esteem scale or the possible selves questionnaires, instead of a global self-esteem scale, indicates that girls, indeed, do identify more strongly with a collective notion of self than do boys. Self-esteem is multidimensional for girls: they report more contradictory or opposing self-attributes (Knox 74). Furthermore, women's self-esteem is more strongly related to social acceptance and inclusion than to accomplishments (Carpenter and Johnson 254). In other words, female self-esteem is more dependent on a collective rather than individual orientation.

Previous research (i.e. Carpenter and Johnson 2001, Knox 1998, and Smith 1999) using the collective self-esteem scale or the possible selves questionnaires, instead of a global self-esteem scale, indicates that girls, indeed, do identify more strongly with a collective notion of self than do boys. They assert that self-esteem is multidimensional for girls: they report more contradictory or opposing self-attributes (Knox 74). Furthermore, women's self-esteem is more strongly related to social acceptance and inclusion than to accomplishments (Carpenter and Johnson 254). In other words, we should consider that female self-esteem can be more dependent on a collective rather than individual orientation.

The current conceptualization of self-esteem is primarily based on an American psychological construct with an emphasis on American, masculinist cultural values, in particular, individualism and independence (Heine et.al. 777). Universalizing tendencies, such as that of self-esteem, are still very prominent in social psychology. Cultural stereotypes about masculinity and femininity impact research on many levels including the language, concepts, and research instruments that are chosen for a study (Lips 97). The majority of research in social psychology has been dominated by men. However, a transformation is underway where many traditional assumptions are being critiqued by critical feminist concerns that recognize past and potential biases (Lips 97-98). This new level of consciousness is a step in the right direction in order to overcome universalizing tendencies in psychological research.

As a result, social psychological theories and measures are being questioned by academics and activists in community psychology and feminist studies. They are identifying the problems caused by traditional approaches that ignore the social context of individual behaviour (Cosgrove and McHugh 818; Hill 761). Further, they are recognizing that psychologists need to work in order to avoid universalizing, which comes from generalizing social phenomena from one context to another without considering the unique environment in which the individual is situated (Bond 589).

In order to overcome universalizing tendencies, social psychologists need to, first, identify that it is a problem. Feminist psychologists argue for the importance of recognizing the diversity of human experience including the influence of cultural background, social class, and gender (Hyde 9). Second, scholars should combine qualitative and quantitative methods in order to determine whether social psychological measures are culturally appropriate. Some feminist scholars advocate qualitative methods, specifically fieldwork and interviews, as a way of overcoming the problem of biased theoretical statistical models and stereotyped research questions (Hyde 9) while others recognize a variety of techniques and promote integration of methods (Campbell and Wasco 773). Although the research process cannot be completely objective and bias-free, scholars should attempt to employ reflexivity - recognizing and acknowledging the existence of biases in their research - and work to decrease them (Hill 759; Lips 85). Much more work on gender needs to be done. Nationally, a gender study (similar to Chen's cultural analysis) could be done to verify whether or not gendered bias exists.[6] Likert scales consisting of various attitudes could be given to girls and to boys in order to determine if boys gravitate toward extreme measures and girls toward midpoint measures, adding to our ability to craft instruments that are truly reflective of the multivalent social factors affecting any study group.

Can there be a universal self-esteem scale? The answer is no; there is a link between response bias and culture that makes it likely that measurement scales are only locally specific. This limits the applicability of comparative quantitative methods from a cross-cultural perspective. As indicated, the theory and measurement surrounding self-esteem is already problematic. Most self-esteem theory reinforces masculine Western values that have an individualist predisposition. Global measurement produces cultural and gender biases. Thus, the use of a universal scale reinforces a narrowly-construed universal notion of the self and the implication that this "self" is the "correct self." This understanding neglects cultural diversity and overlooks gender socialization.

New research is exploring the self as a reflection of culture and of gender and is questioning the embedded Western ideology that is biased toward masculine culture. The multifaceted concept of identity is important when examining self across cultures precisely because self, and the salience of cultural values, varies ethnically and globally. This inclusion bridges the micro-macro gap while addressing cultured and gendered values that may influence self-esteem.