Journal of Youth and Adolescence, Vol. 26, No. 2, 1997

Ethnic and American Identity as Predictors of Self-Esteem Among African American, Latino, and White Adolescents Jean S. Phinney,1 Cindy Lou Cantu,2 and Dawn A. Kurtz3 Received March 18, 1996; accepted May 18, 1996

In order to examine ethnic and American identity as predictors of self-esteem among adolescents, we surveyed 669 American-born high school students (372 Latinos, 232 African Americans, and 65 Whites). Participants completed measures of self-esteem, ethnic identity, American identity, attitudes toward other groups, and demographic variables. Multiple regression analyses of self-esteem were carried out separately for each ethnic group, using ethnic and American identity, othergroup attitudes, gender, age, socioeconomic status, and grade point average as predictors. Results indicated that for all groups ethnic identity was a significant predictor of self-esteem. For the White students only, American identity was a strong predictor of self-esteem and was highly correlated with ethnic identity. Other predictors varied across ethnic groups. Although ethnic identity was a significant predictor of self-esteem, it accounted for a relatively small proportion of the variance, suggesting the importance of other influences on self-esteem.

INTRODUCTION Self-esteem is widely acknowledged to be an important factor in adolescent development. It has been found to be positively associated with genThis research was supported in part by Grant S06 GM-08101 from the Minority Biomedical Research Support Program of the National Institutes of Health. 1Professor, Psychology, California State University, Los Angeles. Received Ph.D. from UCLA. Research centers on the role of ethnicity in development, with a focus on ethnic identity and its implications for adolescents. 2Graduate student, Psychology, California State University, Los Angeles. Research interests include perception and cognition, with an emphasis in physiology. 3Graduate student, Psychology, University of Southern California. Research interests include child abuse and sexual abuse, particularly ethnic issues in abuse.

165 0047-2891/97/0400-0165$12.50/0 C 1997 Plenum Publishing Corporation

166

Phinney et al.

eral psychological well-being, and negatively correlated with depression and hopelessness among adolescents (Harter, 1993). Because of the psychological importance of self-esteem, a considerable amount of research has been devoted to understanding its sources (Baumeister, 1993; Rosenberg, 1986; Wylie, 1979). Research on self-esteem has for the most part focused on individual competence and interpersonal experiences as the bases of selfesteem (Harter, 1993). Although individual sources of self-esteem are clearly important, such a focus ignores the impact of group membership on self-esteem. The purpose of the present study was to examine group identity, specifically ethnic and American identity, as predictors of self-esteem among African American, Latino, and White adolescents. The focus of the study is on global self-esteem, defined by Harter (1993) as "the level of global regard that one has for the self as a person" (p. 88). Self-esteem can be understood in terms of separate domains, and research by Harter (1993) suggests that competence in domains considered important to the individual is the basis for global self-esteem. Most research has focused on domains such as academic ability, athletic competence, social acceptance, or physical appearance. In addition, there is evidence that ethnicity is an important domain for ethnic minority youth (Phinney and Alipuria, 1990). Insofar as group membership is a domain important to adolescents, we would expect that group identity, that is, a positive sense of belonging to one's group, would contribute to self-esteem. The theoretical basis for a link between group membership and selfesteem derives largely from social identity theory (Tajfel & Turner, 1986). Social identity, according to Tajfel and Turner, "consists of those aspects of an individual's self-image that derive from the social categories to which he [sic] perceives himself as belonging" (p. 16). These authors suggest that group members differentiate their own group from other groups and evaluate their own group more favorably as a means of enhancing their self-image. One's identity as a member of a group is seen as an important source of global self-esteem. Research has provided some support for the link between group identity and self-esteem (Crocker et al., 1994). However, the implications of social identity theory are ambiguous for members of ethnic minority groups; according to Tajfel (1981), if their group is viewed negatively by society, they may view themselves negatively. This idea is in accord with early writings on self-esteem among ethnic group members, which assumed that membership in a disadvantaged or lower status group would result in lower self-esteem. Studies showing that Black children often preferred White dolls (Clark and Clark, 1947) were interpreted as evidence for lower selfesteem in a stigmatized group. Similarly, it was assumed in early writings

Ethnic and American Identity and Self-Esteem

167

that Mexican Americans would view themselves negatively because of negative stereotypes assigned to them by the dominant group (Ockerman, 1979). Yet the assumption that lower social status or negative stereotypes would result in lower self-esteem for American ethnic groups has been consistently refuted in empirical research. There is strong evidence that African American adolescents either do not differ from, or score higher than White adolescents on measures of self-esteem (Bowler et al., 1986; Crocker and Major, 1989; Hughes and Demo, 1989; Porter and Washington, 1993; Richman et al., 1985; Rosenberg and Simmons, 1972; Tashakkori, 1992). Similarly, Latino adolescents generally show self-esteem scores equal to or higher than those of other groups (Bowler et al., 1986; Jensen et al., 1982; Martinez and Dukes, 1991; Ockerman, 1979). These empirical findings demonstrate that membership in a lower status ethnic group does not account for high or low self-esteem. Rather than membership per se, the importance of the group membership and feelings about the group appear to be the important factors. Group membership is not a categorical, "all-or-nothing" phenomenon (Turner et al., 1994), and members of ethnic and racial groups vary widely in their sense of belonging to a group and in their evaluation of the group. A positive sense of belonging to one's group should enhance self-esteem, whereas negative attitudes and feelings about one's group may reduce self-esteem (Crocker et al., 1994; Helms, 1990; Phinney, 1993). Nevertheless, there are a number of limitations in existing research on group identity and global self-esteem. First, because of historical and cultural differences among American ethnic groups, the relationship of group identity to self-esteem may differ across groups. More generally, the types of self-beliefs underlying self-esteem differ by ethnicity (Tashakkori, 1992), and the factors underlying self-esteem differ across cultures (Diener and Diener, 1995). Most of the research on ethnic group identity as a factor in self-esteem has focused on African Americans, often in contrast to Whites; there has been little research on this topic with Latino adolescents. Furthermore, although ethnic identity has been widely studied in relation to self-esteem, there has been little consideration of membership in a national group—that is, identity as an American, as a factor in self-esteem. Finally, most studies have looked at simple correlations between various measures of group identity and self-esteem, without considering other factors that may intervene in or confound the relationship. Thus it is not clear how much group identity contributes to self-esteem in comparison to other factors. The present study sought to go beyond previous research by (1) examining predictors of global self-esteem separately among adolescents from three ethnic groups, African American, Latino, and White; (2) extending the concept of group identity to include American identity; and

168

Phinney et al.

(3) including a number of additional variables in order to determine the relative contribution of group identity and other variables to global selfesteem. Research on the role of group identity in self-esteem has focused largely on ethnic identity. Ethnic identity has been conceptualized as a complex construct including a commitment and sense of belonging to the group, positive evaluation of the group, interest in and knowledge about the group, and involvement in social activities of the group (Phinney, 1990). For members of ethnic groups, ethnic identity has been shown to be a highly salient component of identity (Aries and Moorehead, 1989; Phinney and Alipuria, 1990). Ethnic identity is of particular importance during adolescence, when young people begin to face the larger developmental issue of identity formation (Erikson, 1968). In addition to identity concerns regarding occupation, ideology, and gender roles, ethnic minority adolescents are confronted with issues related to being a member of a group that may differ culturally from the mainstream society, that is typically of lower status and power in society, and that may be subject to discrimination. Research suggests that these youth engage in a process of ethnic identity development to explore these issues; this process, over time, leads to an achieved ethnic identity characterized by a strong and secure sense of self as a member of an ethnic group and an accompanying positive self-concept (Phinney, 1989; 1993; Phinney and Chavira, 1992). The view that a strong ethnic identity is positively associated with self-esteem is supported in a number of studies (Belgrave et al., 1994; Phinney, 1992; Wright, 1985). A recent meta-analysis showed a small but consistent relationship between ethnic identity and selfesteem (Bat-Chava and Steen, 1995). We therefore expected that ethnic identity would be one of the predictors of global self-esteem in the current study. However, ethnic identity is not the only group identity for members of a heterogeneous society. American-born adolescents, regardless of their ethnicity, also have a national identity as Americans. In an extensive review article, Scheibe (1983) explored the psychological implications of national identity. He pointed out that one's nation plays an important role in the psychological task of self-definition and that psychologists need to recognize and study it as an aspect of social identity. Theoretically, in line with social identity theory (Tajfel and Turner, 1986), we might expect that a strong feeling of belonging to a powerful nation such as America would contribute to a positive self-concept. However, such an effect would be expected only insofar as American identity is important to the individual. This identity may vary in terms of the extent to which individuals feel their group is included in the concept of America. This idea clearly applies to White Americans, but may be problematic for

Ethnic and American Identity and Self-Esteem

169

ethnic minorities. A recent study suggests that feelings of belonging to America are highly variable among ethnic minority adolescents and are unrelated to self-concept. Phinney and Devich-Navarro (1997) interviewed 52 African American and 46 Mexican American high school students regarding their sense of being American and also gave them measures of American identity (typical item: "I am proud of being American") and of self-concept (using the Tennessee Self-Concept Scale). Although all the adolescents were American born, their interview responses expressed a wide range of views about being American, from a strong identification with America to a denial of being American at all. Furthermore, the questionnaire measure of American identity was unrelated to self-concept. In contrast, a study dealing with Native American adolescents (Oetting and Beauvais, 1991) reported that identification with "the White-American way of life" was positively associated with self-esteem. We know of no studies that have looked at this relationship in White Americans. To explore the relationship between American identity and self-esteem in the present study we assessed how strongly adolescents thought of themselves as American. Various meanings can be attached to the term "American" (Larsen et al., 1992) and these vary within and across ethnic groups (Phinney and Devich-Navarro, 1997). We wished to allow respondents to interpret for themselves what being American meant; therefore, we did not define the term explicitly. Because of the paucity of research, we did not make specific predictions about the relationship of American identity to self-esteem. In addition to identification with one's own group, attitudes toward other groups in a diverse society may be related to how one views oneself. In the study described earlier (Phinney and Devich-Navarro, 1997). attitudes of African American and Mexican American adolescents toward other groups were found to be strongly related to self-concept. Correlations between self-concept and other-group attitudes were .72 for the Mexican American adolescents and .48 for the African Americans; both correlations were significant. More generally, the acculturation literature (e.g., Berry, 1990) has explored the relationship between psychological well-being and the ways in which members of ethnic groups relate to other groups and to the larger society. Acculturation attitudes that favor positive relationships with other groups as well as maintenance of one's own culture, that is, attitudes that support integration, are related to better adjustment (Berry and Kim, 1988). Insofar as self-esteem is one indicator of general well-being, it may be related to attitudes toward other groups, at least for individuals from immigrant backgrounds. A number of other factors have been found to predict self-esteem and therefore need to be taken into account in attempting to understand the

170

Phinney et al.

unique contribution of group identity to self-esteem. We included gender, age, socioeconomic status, and academic achievement in the present study, on the basis of prior research suggesting that they are important to selfesteem across a number of ethnic groups. Research with adolescents has generally found that females have lower self-esteem than males (Harter, 1993). However, this finding varies across ethnic groups. With a very large sample of Hispanic, Black, Asian American, Native American, and White junior and senior high school students, Dukes and Martinez (1994) found that self-esteem differed by both ethnicity and gender. When subgroups were created by ethnicity and gender, Black and Hispanic males had the highest self-esteem, and males from each group scored higher than their female counterparts. However, Black females showed relatively high self-esteem. In a study of Black and White high school students, Richman et al. (1985) reported an interaction of gender and ethnicity; White females were significantly lower in self-esteem than White males and Blacks of both genders, but Black males and females did not differ in self-esteem. We therefore expected gender to predict selfesteem in the White samples, but not in the Black sample. Because of the limited research, no prediction was made for Latinos. The relationship of age to self-esteem during adolescence is unclear. A two-year longitudinal study of three cohorts of Black and White adolescents showed clear increases in self-esteem each year for 7th, 9th, and 11th graders (McCarthy and Hoge, 1982). Another study with adolescents (GrayLittle and Appelbaum, 1979) reported that Black and White tenth graders scored higher than their seventh-grade peers on the Tennessee Self-Concept Scale. However, an early review did not find age differences in selfesteem (Wylie, 1979), and Harter (1993) reported a slight decline in self-esteem for girls during adolescence. Because of the discrepant findings regarding age, we included it in our analyses but did not make a specific prediction about its effect. Socioeconomic status (SES) may also play a role in self-esteem, but the findings vary across studies. Fu et al. (1983) found that Black, Latino, and White preadolescents from lower socioeconomic groups had lower selfesteem than those from higher SES backgrounds. Similarly, using the PiersHarris scale with Black and White high school students, Richman et al. (1985) found that low SES students had lower self-esteem scores than middle or high SES students. A study by Hughes and Demo (1989) with Black adults found a low but significant correlation between social class and selfesteem. However, Ockerman (1979) found no relationship between self-esteem and social class in a sample of African American, Hispanic, and White adolescents. SES was included in the present study but no predictions were made.

Ethnic and American Identity and Self-Esteem

171

Finally, academic achievement has been suggested as an important predictor of self-esteem. Harter (1993) proposed that an important contributor to self-esteem in adolescence is the individual's level of competence in domains considered important. Insofar as academic achievement is considered important to adolescents, it should contribute to their self-esteem. Gray-Little and Appelbaum (1979) reported that academic achievement and self-esteem were positively correlated for Black and White adolescents. In a study of African American and White eighth and tenth graders, Osborne (1995) found a significant positive correlation between grade point average and self-esteem for each gender and racial group at both grade levels except for the tenth-grade African American males. In the present study, we expected that a higher grade point average would be associated with higher self-esteem. In summary, the purpose of this study was to examine ethnic and American identity as predictors of self-esteem in African American, Latino, and White adolescents. Although our focus was on group identity, we also assessed a number of other factors that may affect self-esteem, including attitudes toward other groups, gender, age, social class, and grade point average. Our goal was not to compare groups on self-esteem, but rather to examine different predictors of self-esteem within each group. METHOD Participants The participants were 669 American-born high school students from ethnically diverse schools. The sample included 232 African Americans (117 males and 115 females), 372 Latinos (157 males and 215 females), and 65 White adolescents (34 males and 31 females). The Latinos were primarily Mexican Americans, with small numbers from Central or South America or the Caribbean. The participants were mostly 15-, 16-, and 17-year-olds (M = 16.1 years; SD = 1.1; range: 14-19). The students attended either ethnically diverse public high schools with a predominantly African American and Latino student body; or private parochial high schools that were either mostly African American or mostly Latino. In all settings, the proportion of White students was small. Because the White participants were a small minority in all the school settings, they cannot be considered typical of White adolescents generally. Reports by the participants indicate diversity in SES. The percentage of participants who reported at least one parent who had some advanced education or worked as a professional was 11% for the Latinos, 37% for

172

Phinney et al.

the African Americans, and 57% for the White adolescents. The percentages of those with at least one parent who had some college education or worked in a skilled job were 41% for the Latinos, 75% for the African Americans, and 40% for the White American students. The comparable percentages of those whose parents were unskilled and had no college education were 48% for Latinos, 13% for African Americans, and 3% for White Americans. Because of the differences in SES, as well as other possible ethnic group differences that were not assessed, comparisons were not made across groups except for descriptive purposes.

Procedures Under federal guidelines, parental consent may be waived if a survey is deemed to constitute less than minimal risk to the participants. The investigators requested and obtained such a waiver from the University Institutional Review Board for the Projection of Human Subjects in Research. In addition, among the schools that participated in the study, about one third used passive consent forms, in which parents were informed about the study in advance and could request that their child not participate. When passive consent forms were used, less than 1% of the parents withdrew their child from participation. A wide range of high schools serving predominantly ethnic minority communities throughout the Los Angeles area were contacted and invited to participate, as part of a larger study of identity and adjustment among minority youth (Phinney et al., 1994). In the schools that participated, research assistants visited randomly selected classrooms and invited students to participate. Questionnaires were distributed to all interested students in the classroom. Students were told that their participation was voluntary, that they could withdraw at any time, and that their responses were completely confidential. Less than 2% declined to take part. Completion of the questionnaire took about 15 min. Questionnaires were subsequently screened, and incomplete or randomly completed forms were discarded (less than 3%). For the present study, a selection was made of all those students from the three ethnic groups of interest who were American bom and reported both parents from the same ethnic group as themselves.

Measures The questionnaire included the following measures.

Ethnic and American Identity and Self-Esteem

Ethnic

173

Self-Identification

Participants were asked to select the appropriate ethnic label for themselves and their parents, from a wide range of choices of American ethnic groups, including "mixed." Only African American, Latino, and White American participants who identified themselves and both parents as members of the same ethnic group were included. Ethnic Identity

Ethnic identity was assessed using the 14-item Multigroup Ethnic Identity Measure (MEIM; Phinney, 1992). The measure has a reported reliability of .81 with high school students and .90 with college students. It assesses three aspects of ethnic identity: (1) ethnic identity achievement, based on exploration and commitment (7 items; e.g., "I have spent time trying to find out more about my own ethnic group, such as its history, traditions and customs" and "I have a clear sense of my own ethnic group and what it means to me"); (2) a sense of belonging to, and attitudes toward, one's ethnic group (5 items; for example, "I have a strong sense of pride in my ethnic group and its accomplishments"); and (3) ethnic behaviors and customs (2 items; e.g., " I participate in the cultural practices of my own group, such as special food, music or customs"). Although the three aspects of ethnic identity measured by the scale are conceptually separate, they are all highly correlated and therefore they were not analyzed separately. Items are rated on a 4-point scale ranging from strongly agree to strongly disagree. An ethnic identity score is derived by summing across the 14 items and obtaining a mean. Scores can range from 1 to 4, indicating very low to very high ethnic identity, respectively. In the present study, reliability analysis yielded an alpha equal to .83. American Identity

This was assessed with a single item that asked, "How strongly do you think of yourself as American?" Responses were on a 4-point scale from (4) Very strongly to (1) Very weakly. Other-Group Attitudes

Attitudes toward other groups were measured using a six item scale (Phinney, 1992), which has a reliability of .71 with high school students.

174

Phinney et al.

Participants rate each item on a 4-point scale ranging from strongly agree (4) to strongly disagree (1). The scale includes both positively and negatively worded items, such as, "I like meeting and getting to know people from ethnic groups other than my own" and "I sometimes feel it would be better if different ethnic groups didn't try to mix together." A score is derived by reversing negative items, summing across the items and obtaining a mean; thus a high score indicates positive attitudes toward other groups. The measure had a reliability of alpha equal to .75 with this sample.

Self-Esteem The 10-item Rosenberg (1986) global self-esteem scale was used. It includes both positively and negatively worded items; for example: "I feel that I have a number of good qualities" and "I feel that I do not have much to be proud of." This scale has been widely used in research and found to be highly reliable. In the present study, the reliability analysis showed alpha equal to .83.

Demographic Information Participants supplied information concerning their gender (1, female; 2, male); age in years; grade point average (GPA), reported on a 4-point scale from 1, D or lower, to 4, A- to A; birthplace; parents' occupation, reported on a 3-point scale from 1, unskilled work, to 3, professional or managerial (see description of Participants); and their own and their parents' ethnicity. Only students who met the following criteria were included: American-bora; between the ages of 14 and 19; either Black/African American, Latino/Hispanic, or White/Anglo American, with both parents from the same ethnic group as themselves. RESULTS

As noted earlier, our goal was not to compare ethnic groups, but to examine predictors of global self-esteem within groups. However, before carrying out within-group analyses, we examined differences in the predictor and outcome variables across the three groups. We first conducted an analysis of covariance of self-esteem for the entire sample, with ethnicity, SES, and gender as factors, and ethnic identity, American identity, othergroup attitudes, GPA, and age as covariates. The results showed a significant effect of ethnicity, F(2, 62) = 10.2, p < .001, and gender, F(l, 626)

Ethnic and American Identity and Self-Esteem

175

= 10.1, p < .01, on self-esteem. There were no effects of SES and no interactions among any of the factors as influences on self-esteem. Paired comparisons revealed that African Americans scored significantly higher in self-esteem than the other two groups; Latino and White adolescents did not differ. In all groups, males scored higher than females. The means for self-esteem are shown in Table I, by ethnicity and gender. In addition, three of the covariates had a significant effect on self-esteem: ethnic identity, t = 7.0, p < .001, American identity, t = 2.1, p < .05, and GPA, t = 4.7, p < .001. To examine ethnic group differences among the predictors of self-esteem, a multivariate analysis of variance of all the predictor variables was carried out, using ethnic group as a factor. The analysis revealed a significant over-all effect for ethnicity, Wilks's lambda = .65, F(2, 321) = 22.4, p < .001. Univariate analyses of variance showed significant ethnic group differences in ethnic identity, F(2, 650) = 33.08, p < .001; American identity, F(2, 650) = 6.49, p < .01; other-group attitudes, F(2, 650) = 15.25, p < .001; SES, F(2, 650) = 84.64, p < .001; and GPA, F(2, 650) = 12.98, p < .001. The means for ethnic identity, American identity, and other-group attitudes are shown in Table I by ethnicity. Age and gender did not differ significantly across ethnic groups. The differences in SES were presented in the description of the participants. The significant ethnic differences in both predictor and outcomes variables reinforce the importance of looking at the relationships to self-esteem separately by ethnic group and not making comparisons across groups. Our next step was therefore to examine the predictors of self-esteem separately within each ethnic group. Regression analyses were carried out for each ethnic group separately, using the following variables as predictors of self-esteem: ethnic identity, American identity, other-group attitudes, gender, age, socioeconomic status, and GPA. The partial correlation coefficient for all predictors are shown in Tables II, III, and IV for African Table I. Means and Standard Deviations for Self-Esteem, Ethnic Identity, American Identity, and Other-Group Attitudes

Self-esteem Males Females Ethnic identify American identity Other-group attitudes

African Americans (n = 232)

Latinos (n = 372)

3.37 (.47)

3.07 (.52) 3.17 (.50) 3.00 (.53) 3.16 (.45) 3.05 (.76) 3.22 (.54)

3.41 (.47)

3.33 (.51) 3.26 (.42) 3.23 (.88) 3.07 (.66)

Whites (n = 65)

3.12 (.57) 3.28 (.47)

2.93 (.63) 2.74 (.60) 3.39 (.74) 3.53 (.58)

176

Phinney et al. Table II. Regression Predicting Self-Esteem and Bivariate Correlations for African Americans

B

Predictor Variables Ethnic identity American identity Other-group attitudes Gender SES GPA Age

t

Beta

.22 .20 .06 .03 -.02 -.02 -.12 -.13 .01 .02 .13 .20 .02 .04 Summary statistics: R2 = .09, F(l, 221) = 2.84a

2.92C .92 -.34 -1.89 .29 2.98* .54

Bivariate Correlations

Ethnic Identity American identity Other-Group attitudes SES GPA Gender

Ethnic American Other-Group Esteem Identity Identity Attitudes

SES

GPA

.17* .03 -.01 .04 .18b -.09

.08 .09

.08

-.14a -.08 .15° .03 .12

.30c .02 .08 .13

.05 .09 .12

< .05. bp < .01. cp < .001. ap

American, Latino, and White adolescents, respectively. Also included in each table are the bivariate correlations among all the variables except for age, which was not significantly correlated with any of the other variables. Across all three groups, ethnic identity was a significant predictor of self-esteem. American identity was a strong and significant predictor for the White adolescents but not for the other two groups. In each case, higher group identity (ethnic or American) was related to higher self-esteem. Other-group attitudes were a significant predictor for Latinos, with more positive attitudes predicting higher self-esteem. Among the demographic variables, gender was a significant predictor for Latino and White adolescents; males were higher than females in self-esteem. GPA predicted selfesteem for the African American and Latino adolescents, with higher grades related to higher self-esteem. Age and socioeconomic status were unrelated to self-esteem in all groups. Both the predictors and the amount of variance accounted for differed across ethnic groups. For the African American adolescents, the two significant predictors, ethnic identity and GPA, accounted for only 8.5% of the variance. For the Latino adolescents, four predictors were significant

Ethnic and American Identity and Self-Esteem

177

Table III. Regression Predicting Self-Esteem and Bivariate Correlations for Latino Adolescents Predictor Variables Ethnic identity American identity Other-group attitudes Gender SES GPA

Age

Beta

B

.28 .24 .03 .04 .10 .11 -.22 -.21 -.09 -.07 .16 .12 .06 .03 Summary statistics: R2 = .16, F(7, 357) = 9.75C

t 4.74C .78 2.10a -4.21C -1.76 3.26C 1.28

Bivariate Correlations Esteem Ethnic identity American identity Other-group attitudes SES GPA Gender

Ethnic American Other-group Attitudes Identity Identity

SES

GPA

-.06 .19c .03 .16* .11°

.05 .02

.06

.27c

.05 .14* .11a .20C -.16b

.13a .13a .13a .08

- .08 .08 .13b

< .05. bp < .01. cp < .001. ap

(ethnic identity, other-group attitudes, gender, and GPA), accounting for 16% of the variance. Among the White students, the ethnic and American identities and gender together accounted for 48% of the variance. The bivariate correlations for each group show that the relationship among the predictor variables was different across groups. In particular, American identity has different correlates in each ethnic group. For the African Americans, American identity was related negatively to ethnic identity and positively to attitudes toward other groups; for the Latinos, American identity was unrelated to ethnic identity and positively related to other-group attitudes; for the Whites, American identity was positively related to ethnic identity and unrelated to attitudes toward other groups. DISCUSSION The results of this study provide strong support for the importance of group identity as a predictor of global self-esteem, even when a number of other variables are taken into consideration. Ethnic identity, assessed as

178

Phinney et al. Table IV. Regression Predicting Self-Esteem and Bivariate Correlations for White Adolescents

B

Beta

t

.22

.22 .48 .03

2.09° 4.36*

-.01

-.24 -.01

-2.33° -.12

.09 .01

.11 .01

Predictor Variables

Ethnic identity American identity Other-group attitudes Gender SES GPA Age

.41 .03 -.27

.31 .96 .18

Summary statistics: R2 = .48, F(7, 62) = 7.27c Bivariate Correlations

Esteem Ethnic identity Amer identity

Other-group attitudes SES GPA

Gender

Ethnic Identity

American Identity

Other-group Attitudes

SES

GPA

.44C

.60c -.12 -.08

.13 -.31a

.27°

-.13 -.13

.01 -.08

-.02 -.17

.08

.05

-.20

-.21

.15

.34* -.02

.11

< .05. bp < .01. cp < .001. ap

a broad construct including sense of belonging, positive attitudes, commitment, and involvement with one's group was a significant predictor of selfesteem for African American, Latino, and White adolescents in predominantly non-White schools. Nevertheless, ethnic identity did not account for a large amount of variance. American identity, assessed simply as a sense of being American, strongly predicted self-esteem for the White adolescents but not for the other two groups. The findings regarding ethnic identity are consistent with other research showing a significant relationship between self-esteem and a positive attachment to one's racial or ethnic group (Bat-Chava and Steen, 1995; Crocker et al., 1994; Phinney, 1992; Wright, 1985). The results show that for adolescents in ethnically diverse settings, the higher their ethnic identity—that is, the more committed and positive they feel about their ethnic group, the higher their self-esteem. When adolescents have negative attitudes or are unclear or uncommitted to their ethnicity, their self-esteem is lower. The strength of the relationship is quite similar across ethnic groups. The findings emphasize the point that group membership per se is not the

Ethnic and American Identity and Self-Esteem

179

critical factor in self-esteem. Contrary to the implications of social identity theory (Tajfel and Turner, 1986), the fact that other people view one's group negatively does not necessarily lead to negative self evaluations. Crocker and her colleagues (1994) found that the public scale of collective self-esteem, that is, one's assessment of how others evaluate one's group, had no relationship to personal self-esteem or to other measures of psychological well-being for African Americans. Ethnic group members may view their own group quite differently from the way others view their group; and it is individuals' own commitment and attitudes about their group, rather than evaluations of the group by others, that influence self-esteem. For the small sample of White adolescents, both ethnic identity and American identity were strong predictors of global self-esteem. Furthermore, for these adolescents, the two group identities were significantly related, and American identity may in fact be seen by White adolescents as an aspect of their ethnic identity. The importance of ethnic and American identity for self-esteem among White adolescents in this study should be considered with reference to the context; that is, the White participants were a small minority in settings that were predominantly non-White. A similar strong relationship between ethnic identity and self-esteem was found in a previous study in which White high school students were a small proportion of the sample (Phinney, 1992). In such a setting ethnicity is likely to be more salient than would be the case in largely White settings (McGuire et al., 1978). These results suggest that the importance of ethnic identity to self-esteem may result in part from being in a minority situation and therefore feeling the need for group solidarity (Breakwell, 1986). These findings should be explored in future research in other settings. Although the settings may account in part for the relationship of American identity to self-esteem for the White adolescents, American identity for these students is also likely to be associated with a sense of belonging to the White majority that is dominant in many contexts in society. The sense of empowerment that comes from identification with this dominant group may provide a source of pride for White adolescents. This interpretation is at odds with the view of America as a highly individualistic culture (Markus and Kitayama, 1991), where group identity would be not expected to play a large role. National identity and patriotism may represent one area where individualism gives way to group pride. In contrast, non-White adolescents may have ambivalent feelings about being American, stemming from uncertainty about what "American" means. Many ethnic adolescents feel that "American" or mainstream culture generally means "White" and thus does not include them (Fordham and Ogbu, 1986; Phinney and Devich-Navarro, 1997). The idea that "American" is associated with others rather than the self is reinforced by

180

Phinney et al.

the correlation that was found between other-group attitudes and American identity; the less positive the attitudes that these adolescents held toward members of other groups, the less strongly they thought of themselves as American, and, conversely, positive other-group attitudes were associated with feeling more American. This relationship was significant for both African American and the Latino youths, suggesting that being American is associated with involvement with other groups, that is, with a diverse society. For the White youths, attitudes toward other groups were not significantly related to feeling American. The implication is that for them being American need not involve others. Because being American is not closely tied to the sense of self for the Black and Latino youths, it is not surprising that American identity makes no contribution to their self-esteem. This finding contrasts with that of Oetting and Beauvais (1991), who found a relationship between White American identification and self-esteem among Native American adolescents. In that study, the group studied was different, and identification was assessed as living according to a particular way of life. In contrast, the present study assessed how American participants felt. A particular life style may be indicative of acculturation to the larger society, but may not necessarily mean that one feels more American. Future research needs to explore the various meanings of being American and their relationship to self-esteem. Attitudes toward other groups made a small contribution to global selfesteem for the Latinos but not for the other groups. Latinos, as members of a largely immigrant group, experience varying degrees of acculturation which may influence self-esteem. Rogler et al. (1991) suggest that low acculturation among Hispanics may be associated with low self-esteem. Gil et al. (1994) found that difficulty in communicating in English contributed to low self-esteem, and more generally, that acculturative stress was related to lower self-esteem among American-born Hispanic boys who were low in acculturation. Adolescents who are less acculturated may have poorer language skills, greater difficulty in communicating outside their own group, and fewer friendships with other group members. Acculturation may thus be an important variable even for American-born youths. In the present study, acculturation was not assessed. However, it is possible that less positive attitudes toward other groups are related to lower levels of acculturation, which in turn are related to lower self-esteem. Acculturation should be explicitly assessed in future studies to examine this possibility. Other factors that were found to predict self-esteem generally support previous research. As expected, gender predicted self-esteem in Latino and White adolescents. In both cases, males consistently had higher self-esteem than females. In contrast, gender was not a significant factor in self-esteem for African American adolescents. The findings generally concur with those

Ethnic and American Identity and Self-Esteem

181

of previous studies (Harter, 1993; Hoelter, 1982; Martinez and Dukes, 1991; Richman et al., 1985). The persistent findings of high self-esteem in African American females remains unexplained and should be explored in future research. Academic achievement, as indicated by self-reported GPA, was a significant predictor of self-esteem for the African American and Latino adolescents, but not for the White students. The findings for these adolescents are in accord with work suggesting that academic achievement can make a contribution to self-esteem when it is seen as an important domain (Harter, 1993; Osborne, 1995). Apparently school performance was important for the African American and Latino adolescents in the present study, who attended school in predominantly non-White settings. In a large sample of African American adults, Hughes and Demo (1989) reported the importance of personal efficacy to Black self-esteem. Good school performance may be seen as an indicator of personal efficacy. The same relationship may hold for Latinos, although there is little research on the topic. The findings for White adolescents are difficult to explain; for this group, group identity and gender clearly predominated over academics as a source of self-esteem. Socioeconomic status was not a predictor of self-esteem for the sample as a whole or for any of the three ethnic groups. Although the bivariate correlations showed a relationship between SES and self-esteem for the Latinos, this effect disappeared in the regression analyses. Previous research showing this variable to be related to self-esteem may have been the result of confounding with other, uncontrolled variables. Similarly, age was not a factor in self-esteem in this sample, perhaps for the same reason. In addition, the range of ages was relatively small, with most participants between 15 and 17. Although group identity and a number of other variables contributed to self-esteem in the groups studied, there were large differences among ethnic groups in the total variance accounted for. Because of the large contribution of American identity to self-esteem for the White adolescents, a large amount of the variance was accounted for in this group. However, in the non-White samples relatively small amounts of variance were accounted for. Among African American adolescents, there are clearly many other contributors to self-esteem not included in the present study. Harter (1993) has proposed that a central factor in adolescent self-esteem is the social support individuals receive from significant others such as family and friends. A number of writers have suggested that for Black students, selfesteem is based on reflected appraisals of parents, friends, and teachers (Rosenberg, 1986; Rosenberg and Simmons, 1972). Luster and McAdoo

182

Phinney et al.

(1995) reported that among Black youths, perceptions of parental support and approval made the largest contribution to self-esteem, more than academic and economic accomplishments. The contribution of parental support to self-esteem in that study was greater than the contribution of ethnic identity in the present research. Hoelter (1982) also suggested that Blacks are more likely than Whites to focus on interpersonal relations as a source of self-esteem. Hughes and Demo (1989), with a large sample of African American adults, found that family and friendship made the largest contribution to personal self-esteem. In their study, Black identity (assessed as the degree of closeness that respondents felt toward other Blacks) and racial self-esteem (overall evaluations of Black people as a group) were less important predictors of personal self-esteem than family and friends. Thus, while group identity remains a contributor to self-esteem for this group, it would be appropriate in future research to examine more closely the personal factors associated with Black self-esteem (Harter, 1993). For the Latinos, like the African Americans, the total variance accounted for was not large, in spite of the fact that four variables predicted self-esteem: ethnic identity, gender, grade point average, and other-group attitudes. Clearly, many other factors remain to be explored with this group. Family relationships may be an important contributor to self-esteem for Latinos. Because of the importance of the family in Latino culture (Sabogal et al., 1987), adolescents may draw more support from this source. Acculturation has already been mentioned as an influence on self-esteem. Factors associated with acculturation, such as language proficiency, educational level, and perceived discrimination may have an impact on self-esteem (Gil et al., 1994) and should be considered in future research with Latinos. A limitation of the present study is the fact that the data are correlational, and the direction of effect cannot be demonstrated. It is possible that high self-esteem contributes to group identity, rather than vice versa. In fact, data from a small longitudinal study suggest that the effects are interactive, with each contributing to the other over time (Phinney and Chavira, 1992). Further longitudinal studies are needed to examine how self-esteem develops over time, and how it influences and is influenced by other factors. In addition, these results were obtained in predominantly non-White settings, so that generalizations cannot be made to other settings where the proportion of ethnic minority students is smaller and where Whites are a clear majority. An additional Limitation of this study is the use of a single item to assess American identity; future research is needed to develop better measures and examine this construct in greater depth. Nevertheless, the results provide valuable insights into factors that underlie global self-esteem. Ethnic identity is clearly one important contributor to adolescents' sense of self, and its impact appears to be fairly

Ethnic and American Identity and Self-Esteem

183

consistent across ethnic groups. However, group identity accounted for a small amount of variance; there are clearly many other factors that influence the development of global self esteem. In the present study, grade point average was an important predictor of self-esteem for the minority youth, and girls showed generally lower self-esteem. In considering interventions, it would therefore make sense not only to encourage ethnic pride as a basis for self-esteem, but also to attempt to enhance personal competence (Harter, 1993), for example by providing academic support that might Improve school performance. In addition, the results suggest that girls, especially Latinas and Whites, could benefit from efforts to enhance their sense of competence. A better understanding of the sources of global selfesteem has the potential to help all adolescents lead more satisfying lives.

REFERENCES Aries, E., and Moorehead, K. (1989). The importance of ethnicity in the development of identity of Black adolescents. Psychological Reports, 65: 75-82. Bat-Chava, Y., and Steen, E. (1995). Ethnic identity and self-esteem: A meta-analytic review. Manuscript submitted for publication. Baumeister, R. (1993). Self-Esteem: The Puzzle of Low Self-Regard. Plenum, New York. Belgrave, F., Cherry, V., Cunningham, D., Walwyn, S., Latlaka-Rennert, K., and Phillips, F. (1994). The influence of Africentric values, self-esteem, and Black identity on drug attitudes among African American fifth graders: A preliminary study. /. Black Psychol. 20: 143-156. Berry, J. (1990). Psychology of acculturation. In Berman, J. (ed.), Cross-Cultural Perspectives: Nebraska Symposium on Motivation. University of Nebraska Press, Lincoln. Berry, J., and Kim, U. (1988). Acculturation and mental health. In Dasen, P., Berry, J., and Sartorius, N. (eds.), Health and Cross-Cultural Psychology: Toward Applications. Sage, Newbury Park, CA. Bowler, R., Rauch, S., and Schwarzer, R. (1986). Self-esteem and interracial attitudes in Black high school students: A comparison with five ethnic groups. Urban Educat. 21: 3-19. Breakwell, G. (1986). Coping with Threatened Identities. Methuen, London. Clark, K., and Clark, M. (1947). Racial identification and preference in Negro preschool children. In Newcomb, T., and Hartley, E. (eds.), Readings in Social Psychology. Holt, New York. Crocker, J., Luhtanen, R., Blaine, B., and Broadnax, S. (1994). Collective self-esteem and psychological well-being among White, Black, and Asian college students. Personal. Social Psychol. Bull. 20: 503-513. Crocker, J., and Major, B. (1989). Social stigma and self-esteem: The self-protective properties of stigma. Psychol. Rev. 96: 608-630. Diener, E., and Diener, M. (1995). Cross-cultural correlates of life satisfaction and self-esteem. J. Personal. Social Psychol. 68: 653-663. Dukes, R., and Martinez, R. (1994). The impact of ethgender on self-esteem among adolescents. Adolescence 29: 105 115. Erikson, E. (1968). Identity: Youth and Crisis. Norton, New York. Fordham, S., and Ogbu, J. (1986). Black students' school success: Coping with the burden of "Acting White." Urban Rev. 18: 176-206. Fu, V., Hinkle, D., and Korslund, M. (1983). A developmental study of ethnic self-concept among preadolescent girls. J. Genet. Psychol. 142: 67-73.

184

Phinney et al.

Gil, A., Vega, W., and Dimas, J. (1994). Acculturative stress and personal adjustment among Hispanic adolescent boys. J. Commun. Psychol. 22: 43-54. Gray-Little, B., and Appelbaum, M. (1979). Instrumentality effects in the assessment of racial differences in self-esteem. J. Personal Social Psychol. 37: 1221-1229. Harter, S. (1993). Causes and consequences of low self-esteem in children and adolescents. In Baumeister, R. (ed.), Self-Esteem: The Puzzle of Low Self-Regard. Plenum, New York. Helms, J. (1990). Black and White Racial Identity: Theory, Research, and Practice. Greenwood, New York. Hoelter, J. (1982). Race differences in selective credulity and self-esteem. Sociol. Quart. 23: 527-537. Hughes, M., and Demo, D. (1989). Self-perceptions of Black Americans: Self-esteem and personal efficacy. American J. Sociol. 95: 132-159. Jensen, G., White, C., and Galliher, J. (1982). Ethnic status and adolescent self-evaluations: An extension of research on minority self-esteem. Social Probl. 30: 226-239. Larsen, K., Killifer, C, Csepeli, G., Krumov, K., Andrejeva, L., Kashlakeva, N., Russinova, Z., and Pordany, L. (1992). National identity: A new look at an old issue. J. Social Behav. Personal. 7: 309-322. Luster, T., and McAdoo, H. (1995). Factors related to self-esteem among African American youths: A secondary analysis of the High/Scope Perry preschool data. J. Research Adolesc. 5: 451-467. Markus, H., and Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psycholo. Rev. 98: 224-253. Martinez, R., and Dukes, R. (1991). Ethnic and gender differences in self-esteem. Youth Soc. 22: 318-338. McCarthy, J., and Hoge, D. (1982). Analysis of age effects in longitudinal studies of adolescent self-esteem. Develop. Psychol. 18: 372-379. McGuire, W., McGuire, C, Child, P., and Fujioka, T. (1978). Salience of ethnicity in the spontaneous self-concept as a function of one's ethnic distinctiveness in the social environment. J. Personal Social Psychol. 33: 743-754. Ockerman, J. (1979). Self-Esteem and Social Anchorage of Adolescent White, Black, and Mexican American Students. R & E Research Associates, Palo Alto, CA. Oetting, E., and Beauvais, F. (1991). Orthogonal cultural identification theory: The cultural identification of minority adolescents. Int. J. Addict. 25: 655-685. Osborne, J. (1995). Academics, self-esteem, and race: A look at the underlying assumptions of the disidentification hypothesis. Personal Social Psychol. Bull. 21: 449-455. Phinney, J. (1989). Stages of ethnic identity development in minority group adolescents. /. Early Adolesc. 9:34-49. Phinney, J. (1990). Ethnic identity in adolescents and adults: A review of research. Psychol. Bull. 108: 499-514. Phinney, J. (1992). The Multigroup Ethnic Identity Measure: A new scale for use with diverse groups. J. Adolesc. Res. 7: 156-176. Phinney, J. (1993). A three-stage model of ethnic identity development. In Knight, M. B. and G. (eds.), Ethnic Identity: Formation and Transmission Among Hispanics and Other Minorities. State University of New York Press, Albany. Phinney, J., and Alipuria, L. (1990). Ethnic identity in college students from four ethnic groups. J. Adolesc. 13: 171-184. Phinney, J., and Chavira, V. (1992). Ethnic identity and self-esteem: An exploratory longitudinal study. J. Adolesc. 15: 271-281. Phinney, J., and Devich-Navarro, M. (1997). Variations in bicultural identification among African American and Mexican American adolescents. /. Res. Adolesc. 7: 3-32. Phinney, J., Dupont, S., Espinosa, C., Revill, J., and Sanders, K. (1994). Ethnic identity and American identification among ethnic minority youths. In Bouvy, A., van de Vijver, F., Boski, P., and Schmitz, P. (eds.), Journeys into Cross-Cultural Psychology. Swets & Zeitlinger, Amsterdam. Porter, J., and Washington, R. (1993). Minority identity and self-esteem. Ann. Rev. Sociol. 19: 139-161.

Ethnic and American Identity and Self-Esteem

185

Richman, C., Clark, M., and Brown, K. (1985). General and specific self-esteem in late adolescent students: Race x gender x SES effects. Adolescence 20: 555-566. Rogler, L., Cortes, D., and Malgady, R. (1991). Acculturation and mental health status among Hispanics. Am. Psychol. 46: 585-597. Rosenberg, M. (1986). Conceiving the Self. Kreiger, Melbourne. Rosenberg, M., and Simmons, R. (1972). Black and White Self-Esteem: The Black Urban Child. American Sociological Association, Washington, DC. Sabogal, F., Marin, G., Otero-Sabogal, R., Marin, B., and Perez-Stable, E. (1987). Hispanic familism and acculturation: What changes and what doesn't? Hisp. J. Behav. Sci. 9: 397412. Scheibe, K. (1983). The psychology of national identity. In Sarbin, T., and Scheibe, K. (eds.), Studies in Social Identity. Praeger, New York. Tajfel, H. (1981). Human Groups and Social Categories. Cambridge University Press, Cambridge, UK. Tajfel, H., and Turner, J. (1986). The social identity theory of intergroup behavior. In Worchel, S., and Austin, W. (eds.), Psychology of Intergroup Relations. Nelson-Hall, Chicago. Tashakkori, A. (1992). Gender, ethnicity, and the structure of self-esteem: An attitudes theory approach. J. Social Psychol. 133: 479-488. Turner, J., Oakes, P., Haslam, S. A., and McGarty, C. (1994). Self and collective: Cognition and social context. Personal. Social Psychol. Bull. 20: 454-463. Wright, B. (1985). The effects of racial self-esteem on the personal self-esteem of Black youth. Int. J. Intercult. Relat. 9: 19-30. Wylie, R. (1979). The Self Concept (Vol. 2). University of Nebraska Press, Lincoln.