The rise of Latino and Asian immigration to the United States during the past two decades (Capps and Fortuny 2006) and the replacement of African-Americans by Latinos as the largest U.S. racial/ethnic minority (U.S. Census Bureau News 2006) have precipitated heightened national interest in the acculturation process and its outcomes among immigrant groups and individuals. Because the host culture is dominant in terms of power, status, and privilege, it is understandable that the acculturation of immigrants to it has been the main focus of U.S. acculturation research. As the editors of this AJCP Special Section point out, much less attention has been paid to the “other side” of mutual acculturation: namely, acculturation processes and outcomes among host groups and individuals who are in contact with immigrants. While the host culture in the U.S. is strongly European-based, host individuals may be of any racial/ethnic group. Using host versus immigrant categories as a grouping variable, without regard for individuals’ race or ethnicity, obscures the diversity of the host group. A foundational concept of this paper is that, in multi-ethnic contexts as well as in ethnic “enclaves,” racial/ethnic minorities who have resided in the host country for one or more generations are part of a host group relative to their newly-arriving counterparts. The ways in which members of these racial/ethnic minority host groups respond to contact with immigrants are important components of a full understanding of the mutual acculturation that occurs between host and immigrant groups.

In this paper, we develop the concept of ethno-generational group and use it as a variable to clarify aspects of acculturation among adolescents in a multicultural school context. Ethno-generational group is a construct that takes into account one’s racial/ethnic group as well as the number of generations one’s family has been in the U.S. The ethnic component consists of one’s self-labeling as a member of an ethnic group. The generational group portion of this variable is based on whether one is U.S.-born as well as whether one’s parents are U.S.-born. The ethno-generational groups on which we focus are European-American host, Latino host, Latino immigrant, Asian-American host, and Asian-American immigrant adolescents.

Although generational group has been used as a marker of acculturation, we agree with Phinney’s assertion (2003) that neither ethnic group membership nor generational group status are psychological variables. We use ethno-generational group as a grouping variable and assess acculturation outcomes directly. By comparing ethno-generational groups, we not only partially redress the relative inattention that has been paid to acculturation-related outcomes for host individuals in contact with immigrants, we also recognize the growing importance of non-European members of the U.S. host culture. Host versus immigrant group comparisons within and among various racial/ethnic groups provide a clearer picture of acculturation processes because they separate generational group differences from ethnic group differences.

Multi-racial classes of ninth graders enrolled in Los Angeles County high schools provide an especially fortunate context in which to study ethno-generational similarities and differences in mutual acculturation. First, according to the 2000 U.S. Census (http://www.census.gov), Los Angeles County is a region in which no racial/ethnic group is in the majority and immigrants make up about 36% of the population. Second, in schools and classrooms that are highly diverse with respect to both race/ethnicity and generational group (i.e., in which the major racial/ethnic groups are nearly equally represented and there is a nearly even split between U.S.-born and immigrant students), the “host” culture is not restricted to U.S.-born European-American students, but includes earlier waves of racial-ethnic minority students. Third, a focus on ninth graders is especially important in the Los Angeles school context because ninth graders are typically drawn from a variety of middle schools, are in their first year in a new school and form new peer relations. The processes that converge during this transition period are likely to promote accelerated acculturation-related change among such students. In the present paper, we consider how ethno-generational group informs our understanding of racial/ethnic groups of adolescents in contact and we provide evidence concerning four acculturation-related variables. In the following sections, we describe the relevance of three theoretical perspectives to our examination of host and immigrant adolescents in a school context.

Acculturation Perspective

Within cultural psychology, “psychological acculturation refers to the process by which individuals change, both by being influenced by contact with another culture and by being participants in the general acculturative changes underway in their own culture” (Berry, 1990, pp. 234–235). The psychological acculturation of new immigrants is a complex phenomenon (e.g., Bourhis et al. 1997) that is influenced by many factors, including achieved and ascribed characteristics. Among these characteristics are culture, skin color and other phenotypic qualities; social, legal, and economic statuses; language, country of origin and (most importantly for the present research) generational group. Berry et al. (1986), Bourhis et al. (1997) and others conceptualize immigrant acculturation as posing a dilemma concerning (1) the value or importance of maintaining one’s cultural identity and characteristics and (2) the value or importance of relationships with other ethnic groups. Berry et al. (1986) represented this dilemma in a two-dimensional framework that allows for individual variation along each of these two dimensions independently. We call these dimensions ethnic identity affirmation/belonging and outgroup orientation, respectively. Berry et al (1986) conceptualization of acculturation has been widely used by others to study immigrant stress and coping as well as adaptation to the host culture. These two dimensions constitute two of the four acculturation-related outcome variables in the present study.

Stated most simply, ethnic identity affirmation/belonging consists of one’s feelings of belonging to an ethnic group and the extent to which one affirms membership in the group (Phinney 1992). It is widely acknowledged to be a very complex component of one’s social identity. Outgroup orientation is conceptualized as the strength of one’s willingness, and the value one attaches to, spending voluntary time with people from groups other than one’s own. The term “outgroup” reflects the tradition in social psychology of labeling all groups other than one’s own as outgroups.

In the U.S. context, one’s ethnic identity affirmation/belonging may differ among the racial/ethnic groups. According to Phinney et al. (2006), research has consistently shown that members of the dominant cultural group (European Americans in the U.S.) are lower in ethnic identity than members of immigrant ethno-cultural groups. However this statement does not clarify the extent to which ethnic group dominance versus generational status contributes to this difference. The present study allows us to do separate tests for ethnic group and generational status differences in ethnic identity affirmation/belonging.

There is some evidence that the strength of both ethnic identity and outgroup orientation vary with generational status. Some of the relevant research has focused on families. For example, Pyke (2005) interviewed grown children in the U.S. (ages 18–26 years) and found younger siblings were more Americanized and less identified with their ethnic group of origin than their older siblings. Verkuyten (2003) showed a decrease in ethnic identity among immigrants to the Netherlands, the longer immigrants stay in the host culture. In addition, Kwak and Berry (2001) found that later generations of Asian heritage adolescents in Canada had higher outgroup orientation than their parents. However, given the present study’s focus on early adolescents of approximately the same age, we have no prior reason to expect differences in initial or final levels of outgroup orientation for host versus immigrant groups.

Identity Development Perspective

The challenges that immigrants face are compounded when the immigrants are adolescents (e.g., Berry et al. 2006; Rumbaut and Portes 2001). Marcia’s (1966) theorizing concerning identity development can be used to illuminate developmental dilemmas faced by host and immigrant adolescents in contact. Marcia proposed a two-dimensional model that juxtaposes degree of ethnic identity commitment and extent of ethnic identity exploration. Ethnic identity commitment in Marcia’s theorizing is conceptually and operationally the same as ethnic identity affirmation/belonging, already described (Phinney and Ong, in press). In contrast, ethnic identity exploration concerns the extent to which one engages in activities to learn about the culture and traditions of one’s ethnic group. The third acculturation-related outcome in the present report is a measure of ethnic identity exploration, for both host and immigrant adolescents.

Most prior research on generational status (including most of Phinney’s own work) did not separate the two ethnic identity components. Instead, they used an overall ethnic identity measure, which is an aggregate of both affirmation/belonging and exploration component. This makes differential hypotheses concerning them difficult to formulate. Nevertheless the relationship of these two components of ethnic identity is of interest because of their importance to Marcia’s (1966) theorizing concerning the resolution of adolescent identity crises (e.g., Phinney and Ong, in press). Furthermore, Whitehead et al. (2008) demonstrated the usefulness of Phinney’s (1992) ethnic identity exploration—ethnic identity affirmation/belonging distinction for explaining adolescents’ warmth toward their own group and toward other racial/ethnic groups. In addition, Phinney (2003) suggested that experiencing discrimination may elicit such exploration, which in turn promotes ethnic identity affirmation/belonging. Insofar as ethnic groups differ in the levels of discrimination they experience, levels of ethnic identity exploration may differ among ethnic groups as well. In the present study, we tested whether the levels of each of the two ethnic identity components differ by ethnic group and generational status, at the beginning and at the end of 10 weeks of intergroup contact in an educational setting.

The fourth acculturation-related outcome measured in the present study is national identity, conceptualized as a sense of belonging to, and affirmation of membership in, a national group. In their comparison of adolescent immigrants in 13 nations, Phinney et al. (2006) showed that the U.S. ranks third highest in the correlation between ethnic identity and national identity (lower than New Zealand and Australia, but higher than Canada). However, Phinney (2003) asserted that both national and ethnic identity affirmation/belonging are assumed to change across time, context, and generations of immigrants. Phinney and her colleagues reported the relationship between national identity and overall ethnic identity among Mexican-American adolescents in two studies. In a 1997 study (Phinney and Devich-Navarro 1997), the correlation between national identity and an aggregate measure of ethnic identity was not significant, but information concerning the generational status of the participants and the magnitude of the correlation was not reported. In more recent research, Phinney et al. (2006) obtained a moderately positive correlation (0.51) between national identity and (aggregate) ethnic identity among Mexican-American immigrants. The measures of American and ethnic identity used in the two studies were similar, though not identical. The latter study also included an examination of the relationship between national identity and ethnic identity as a function of generational status for several thousand adolescents in 13 countries, with ethnicity unspecified. The authors found that as length of residency in the host country increased, national identity increased. Furthermore, the strength of the relationship between national identity and ethnic identity increased with length of residency.

Sidanius and Pratto (1999) suggested that, as social status decreases, the strength of the relationship of ethnic identity variables to one’s feelings of being American also decreases. Schlesinger (1998) and Renshon (2005) suggested that these two identities may even be somewhat incompatible for some groups. Staerklé et al. (2005) and Weisskirch (2005) found that high status groups show more positive association between ethnic identity and American identity than their low status counterparts, for whom this relationship is sometimes negative. This pattern of results has been called the group status asymmetry effect. A specific form of this effect was obtained by Staerklé et al. (2005), who compared adults in 11 countries and found an ethnic “majority-minority asymmetry.” Specifically, there was a stronger relationship between national identification (assessed via psychological closeness to respondents’ country) and ethnic identification (assessed via perceived closeness to one’s ethnic group) in the majority group, as compared to minority group members.

Because we focus exclusively on adolescents in the U.S. in the present study, we conceptualize and operationalize the national identity construct as American identity. Based primarily on the results of Phinney and colleagues’ studies with adolescents summarized above, we propose several hypotheses concerning generational status differences in the relationship between national identity and one or both of the components of ethnic identity. Specifically, prior to taking the multicultural education class, host groups of adolescents will show (1) higher levels of national identity (Phinney et al. 2006) and (2) a stronger correlation between national and ethnic identity than their immigrant adolescent counterparts (Phinney et al. 2006; Staerkle et al., 2005). Hypotheses concerning these variables following the 10 weeks of intergroup contact are more difficult to specify because, to our knowledge, changes in levels of national identity, as well as in the relationship of national identity to ethnic identity following intergroup contact have not been previously studied.

Group Dominance Perspective

In the U.S. context, racial/ethnic groups differ in status within the social hierarchy. For example, the relative status of the three racial/ethnic groups in our research differs, both within American society, as shown by status ratings (Sidanius and Pratto 1999, pp. 52–53), and within Los Angeles County, as measured by family income (U.S. Census Bureau 2003), with European Americans higher than Asian Americans, who are higher than Latinos. However, these ethnic group considerations do not point toward a single a priori hypothesis concerning group differences on the variables under study. This is because an ethnic group’s social status within a given context (such as the school) may be inconsistent with its degree of dominance, in terms of power and authority over others, in the larger culture.

The situation of Asian Americans illustrates this distinction. While Asian-American adolescents may have relatively high status in a school context, they remain a non-dominant group, relative to European Americans, in the larger society (Oyserman and Sakamoto 1997). A theory about social status in the school context would suggest that Asian Americans would view themselves, and may be viewed by others, as more similar to European Americans (and different from Latinos) on a number of acculturation-related outcomes. For example, Latino students may view their Asian-American classmates as a “model minority,” (see Oyserman and Sakamoto 1997) which is more similar to European-American students than to themselves, which may result in being less willing to interact with members of either of the other two groups. Similarly, in the school setting, Asian-American students may perceive themselves as comparable in status to European-American students and distance themselves from Latinos. On the other hand, a theory about group dominance in society would suggest that Asian Americans would view themselves (and be perceived by others) as a culturally marginalized ethnic group along with Latinos (and different from European Americans). In light of the complexity and number of these possible ethnic group differences, we employ ethno-generational status as a grouping variable in an exploratory manner, allowing for the emergence of group differences without making specific predictions about the directions they will take, or whether they will emerge at all.

Intergroup Contact Setting

The intergroup contact setting for the present study is a mandatory high school multicultural education course. Such classes provide a favorable setting for studying acculturation in early- versus later-generations of immigrants. Our knowledge of the multicultural education literature suggests that ethnic identity exploration will be higher following the coursework. There is also prior research indirectly suggesting that such classes may have differential effects for these two groups, at least among college students. For example, Arends-Toth and van de Vijer (2003), Berry and Kalin (1995), and Verkuyten and Thijs (2002) found that multicultural educational ideologies may threaten the ethnic identity of majority groups. However, we know of no such research with adolescents and there is little direct evidence concerning generational differences in relationships among acculturation variables. Most of the prior studies of multicultural education make comparisons between “majority” versus “minority” groups, without regard for generational status.

Overview of the Present Study

The present research is a comparison of host and immigrant adolescents (separately by racial/ethnic group membership) with respect to four acculturation-related outcomes. These outcomes are: ethnic identity affirmation/belonging, ethnic identity exploration, outgroup orientation, and American identity. We conceptualize earlier generations of ninth graders from three racial/ethnic groups as part of the host culture to later-arriving generations of ninth graders. First, we use a MANOVA to test for differences among the five ethno-generational groups at the beginning of a mandatory ninth grade cultural awareness class. Next, we repeat the MANOVA on Time 2 data. Following each MANOVA, we conduct four planned comparisons. The first two comparisons test for differences between host and immigrant groups, separately for Asian-American and Latino adolescents. The third comparison tests for the pattern derived from the school social status perspective, while the fourth comparison tests a hypothesis derived from the group dominance in society perspective. Then we report exploratory t-tests of change on the acculturation-related variables within each ethno-generational group. Finally, for each of the five ethno-generational groups, we provide a pair-wise correlation matrix of all four acculturation-related variables measured at Time 1 and again at Time 2. However, we are concerned with only those correlations relating to national identity and the two ethnic identity components.

Method

Participants

Respondents were enrolled in mandatory Life Skills classes in a Los Angeles County public school. Data were collected during Fall 2003, Spring 2004, Fall 2004, and Spring 2005. A total of 24 class sections were surveyed (6 sections per term). The participants’ racial/ethnic categories are based on their self-report to an item that allowed them to choose as many groups as “applies to you” from a list of eight alternatives, including “Other (write in) _______,” which allowed for writing in a specific group name not listed. For the present study, only those who chose a single listed category were included. Thus, subgroups of Asian Americans or of Latinos cannot be distinguished, reflecting the choice made by each respondent to self-categorize using a broad term rather than writing in a more specific one.

After data screening and statistically combining respondents across semesters, the sample size is 313, consisting of 153 (48.9%) males and 160 (51.1%) females, with a mean age of 14.17 years. European Americans constituted 32.3%, Latinos 35.1% and Asian Americans 32.6% of the sample. Of the 313 participants, 189 (60.4%) are members of the host group and 124 (39.6%) are members of the immigrant group. The percentages of the five ethno-generational groups are: 32.3% European-American host, 18.2% Latino host, 16.9% Latino immigrant, 9.9% Asian-American host, and 22.7% Asian-American immigrant. The numbers of immigrant European-Americans and of African-Americans were too small to form ethno-generational groups for analysis.

Each student was enrolled in one of 24 sections of the course, each consisting of 30–40 students. For each section, the racial/ethnic makeup of the enrolled students was calculated and averaged across the 24 sections. A chi-square test confirmed that racial/ethnic membership of the students in each of the 24 sections did not deviate significantly from the overall sample.

Procedure

The mandatory multicultural education class we studied was designed, in part, to promote intergroup tolerance. The 24 sections met 5 days/week, 1 h/day. In 16 of the sections, a credentialed teacher taught each section 5 days/week during weeks 1 and 10. During weeks 2 through 9, each of these 16 sections was taught by its credentialed teacher 4 days/week and led by a pair of college student facilitators 1 day/week (under direct supervision of the credentialed teacher). The college student facilitator pairs led discussions and interactive small group exercises on stereotyping, prejudice and discrimination (e.g., racism, sexism, and homophobia) as well as on strategies for preventing conflict and promoting tolerance, following a curriculum (called STOP, Students Take Out Prejudice). All facilitators were enrolled in a college class that gave credit for their eight-week internship. The remaining eight sections were taught exclusively by a credentialed teacher (5 days/week for 10 weeks) who delivered aspects of the STOP curriculum 1 day/week. In all 24 class sections, surveys were conducted in weeks 1 and 10 of the 10-week course (Time 1 and Time 2, respectively). All data were collected in accord with procedures specifically approved by a university institutional review board. Initial data analyses showed no differences on the assessed variables between the facilitator-assisted sections and those entirely taught by a credentialed teacher. Data from all 24 sections of the class were statistically combined for analysis.

Measures

Generational Status

This variable was constructed from students’ responses to two items: whether they were born in the U.S., and the number of their parents born in the U.S. Each host generational group consists of ninth graders who were born in the U.S. and have both parents born in the U.S. Each immigrant generational group consists of ninth graders who were either (1) born outside the U.S. or (2) if U.S. born, have atmost one parent born in the U.S. Thus, in each immigrant group, either the students themselves or at least one of their parents is foreign-born. This definition reflects our belief that having one immigrant parent connects young adolescents to the country of origin to an important extent. This definition is also in accord with Rumbaut (2004) who makes a conceptual distinction based on whether one or both parents are foreign born and provides data to support it.

The remaining variables reported in this study were assessed via 7-point Likert scales, anchored by the terms “strongly disagree,” “disagree,” “somewhat disagree,” “neither agree nor disagree,” “somewhat agree,” “agree,” and “strongly agree.”

Acculturation-Related Variables

Items from Phinney’s (1992) Multigroup Ethnic Identity Measure (MEIM) were used to assess three of the four acculturation-related constructs: (a) five items for ethnic identity affirmation/belonging (α = 0.83), (b) seven items for ethnic identity exploration (α = 0.82) and (c) four items for outgroup orientation (α = 0.92). Rudmin and Ahmadzadeh (2001) recommend this method, whereby each of the acculturation constructs is measured directly because it avoids the ambiguity inherent in using compound items that assess assimilation, marginalization, separation, and integration, which comprise the intersection of ethnic identity affirmation/belonging and outgroup orientation. Means for each construct were calculated to form composite scale scores. The fourth acculturation-related outcome (American identity) was assessed using Phinney and Devich-Navarro (1997) American identity questionnaire (AIQ), except that we replaced their item about “mainstream American culture” with one concerning “attachment toward this country.” A principal components factor analysis with varimax rotation confirmed the unitary factor structure of this seven-item scale and internal consistency was good (α = 0.82).

Results

Ethno-Generational Differences at Time 1

A MANOVA (Tabachnick and Fidell 2001, p. 322) was used to analyze differences among the five ethno-generational groups on the four acculturation-related variables assessed prior to the multicultural education class. Tests for outliers, normality, homogeneity of variance-covariance matrices, and linearity were run and were shown to be within acceptable ranges. MANOVA results showed that all four outcomes significantly differed among the five ethno-generational groups as follows: ethnic identity exploration (F(4,313) = 2.364, p = 0.053), ethnic identity affirmation/belonging (F(4,313) = 6.950, p = 0.001), outgroup orientation (F(4,313) = 2.864, p = 0.024), and American identity (F(4,313) = 8.917, p = 0.001).

Following the MANOVA using all five ethno-generational groups, a series of one-way ANOVAs was run to test for Asian-American host versus immigrant group differences on the four outcomes. Another series of one-way ANOVAs was run to test for Latino host versus immigrant group differences on the four outcomes. No significant results were found among these eight ANOVAs (see Table 1 for details). F-tests on each of the outcomes for which significant differences had been found in the MANOVA were run to detect patterns associated with the social status in-the-school versus group dominance-in-society perspectives. Results are shown in Table 2. At Time 1, results for ethnic identity affirmation/belonging support the former perspective; results for outgroup orientation and American identity support the latter perspective; and results for ethnic identity exploration support neither perspective.

Table 1 Immigrant and host comparisons for Asian-American and Latino Groups at Time 1 and Time 2
Full size table
Table 2 Host group comparisons at Time 1 and Time 2
Full size table

Ethno-Generational Differences at Time 2

A MANOVA was performed on all four outcome variables at Time 2, again using the five ethno-generational categories as the grouping variable. Tests for outliers, normality, homogeneity of variance-covariance matrices, and linearity were run and were shown to be within acceptable ranges. Results showed that ethnic identity affirmation/belonging (F(4,313) = 3.437, p = 0.009) and American identity significantly differed (F(4,313) = 6.417, p = 0.001) among the five ethno-generational groups. Results for ethnic identity exploration (F(4,313) = 2.165, p = 0.073) and outgroup orientation (F(4,313) = 2.085, p = 0.083) showed a trend toward differences among the five groups.

Following the above MANOVA, a series of one-way ANOVAs was run to test for Asian-American host versus immigrant group differences on the four outcomes and another series of one-way ANOVAs was run to test for Latino host versus immigrant group differences on the four outcomes. Consistent with the ANOVAs on Time 1 data, no significant differences were found among these eight ANOVAs (see Table 1 for details). F-tests on each of the outcomes for which significant differences had been found in the MANOVA were run to test for the social status-in-the-school versus group dominance-in-society patterns (Table 2). At Time 2, results for outgroup orientation and American identity support the latter perspective and results for the two ethnic identity constructs support neither perspective.

T-Tests for Changes in Means on Acculturation-Related Variables

In the absence of specific hypotheses concerning changes on the four acculturation-related variables from Time 1 to Time 2, exploratory t-tests were run separately for the five ethno-generational groups. Results of this series of paired-sample t-tests on ethnic identity exploration, ethnic identity affirmation/belonging, outgroup orientation, and American identity scores, at Time 1 and again at Time 2, are shown in Table 3. Seven of the 20 comparisons revealed differences from Time 1 to Time 2. A Bonferroni correction requires that each comparison be significant at p < 0.0025 to be significant at the 0.05 level overall. Using this criterion, ethnic identity exploration among Asian-American immigrants and European-American host students significantly increased from Time 1 to Time 2.

Table 3 Time 1- versus Time 2 differences in acculturation-related variables
Full size table

Correlations Between Pairs of Acculturation-Related Variables

Table 4 presents the bivariate correlations for the four acculturation-related variables at Time 1 and Time 2 for the five ethno-generational groups. Results showed that, at the beginning of the multicultural education class, ethnic identity exploration, and ethnic identity affirmation/belonging were differentially related to American identity for the various ethno-generational groups. Specifically, for the European-American host group at Time 1, both components of ethnic identity were related to American identity. For the Asian-American host group at Time 1, only the ethnic identity affirmation/belonging was related to American identity; and for the Latino host group at Time 1 only exploration of one’s ethnic identity was related to American identity. For the immigrant groups, neither of the ethnic identity variables was related to American identity. The same pattern of relationships was obtained at the end of the ten-week intergroup contact.

Table 4 Correlations of ethnic identity exploration and ethnic identity affirmation/belonging with American identity for each ethno-generational group at Time 1 and Time 2
Full size table

Z-tests of difference between Asian-American host and immigrant groups and between Latino host and immigrant groups were run on those pair-wise correlations reported above for which one of the pair was significant, to test directly whether one correlation was significantly higher than the other member of the pair. Results were as follows: Time 1 Asian-American host versus immigrants for American identity and ethnic identity affirmation/belonging, z(102) = −1.651, p = 0.049; Time 1 Latino host versus immigrants for American identity and ethnic identity exploration, z(110) = −1.915, p = 0.027; Time 2 Asian-American host versus immigrants for American identity and ethnic identity affirmation/belonging, z(102) = −1.288, p = 0.098; and Time 2 Latino host versus immigrants for American identity and ethnic identity exploration, z(110) = 1.111, p = 0.133. Results of the z-tests a Time 1, but not at Time 2, supported the hypothesis that the host group correlations between American identity and ethnic identity components would be higher than the corresponding correlations for the immigrant groups.

Discussion

We believe we are among the first to use ethno-generational membership as a grouping variable to examine acculturation-related constructs. Our analyses demonstrate the importance of taking both ethnic group membership and generational status into account. We pay special attention to results for the European-American, Asian-American and Latino host groups, for which we developed several theory-based hypotheses.

First, we consider the results of the MANOVAs at Time 1 and Time 2 and their respective follow-up F-tests. Although there were no mean differences between the host and immigrant groups on any of the four acculturation-related outcomes, either at Time 1 or at Time 2, this overall finding masks interesting differences. In order to detect patterns related to the group dominance-in-society perspective, ethnic group membership was used as the grouping variable, while holding generational status constant. Results for ethnic identity exploration were not consistent with this perspective, given that there were no differences among the three ethnic groups at either Time 1 or Time 2. The pattern consistent with the group status-in-the-school hypothesis (namely that Asian-American and European-American adolescents are more similar to each other than either group is to Latinos) was obtained for only one of the outcomes (i.e., ethnic identity affirmation/belonging) and only at Time 1.

In contrast, the results for two of the four outcomes were consistent with a group dominance-in-society perspective at both Time 1 and Time 2. Specifically, for both outgroup orientation and American identity, both prior to and following the multicultural education course, Asian-American and Latino host adolescents were more similar to each other than either group was to the European-American host group (which showed higher levels of both outcomes). The European-American host adolescents’ higher mean on outgroup orientation may be taken as evidence that they are relatively more disposed toward interacting with racial/ethnic minorities. Another possibility is that European-American host adolescents are more likely than their minority group counterparts to report socially desirable responses. For example, Judd et al. (1995) suggested that social desirability is more likely to occur among majority group members than among their minority group counterparts, especially when they are surveyed together. The European-American host students’ higher level of American identity than their minority host group counterparts is also open to several interpretations. First, it may be due to it being easier for European Americans to identify as “Americans” because the concept of being an American is infused with more of the traits associated with their ethnic group and fewer of the traits associated with racial/ethnic minorities (Mummendey and Wenzel 1999; Sidanius et al. 1997). In a related vein, it may be a result of minority host group members experiencing obstacles to being accepted by host groups as true Americans (e.g., Rumbaut and Portes 2001).

We now turn to a discussion of the results of the t-tests of mean differences across time periods. Both the European-American host group and the Asian-American immigrant group showed a significant increase in ethnic identity exploration from week 1 to 10. Because these two changes across time were obtained as part of an exploratory series of twenty t-tests, rather than emerging in the context of a priori hypothesis testing, their reliability needs to be established in subsequent research. There are at least two plausible reasons for the emergence of significant increases in ethnic identity exploration for these two particular groups. In the present sample, these two groups are moderate in absolute size, but largest in number, so there is relatively greater power to detect a change in exploration for these groups, relative to the others. Second, these two groups constitute the largest proportions in the classrooms we studied. Thus, all other things being equal, the opportunities for students in each of these two groups to talk with other members of their group about the implications of being a member of their group are higher relative to the other three ethno-generational groups.

Next, we discuss those pair-wise correlations for which we had a priori hypotheses. At both Time 1 and Time 2, all three host groups showed a significant relationship between American identity and one or both ethnic identity components, while none of the immigrant groups did. Furthermore at Time 1, z-tests showed that these correlations for host versus immigrant groups within race/ethnicity were significantly different, in accord with the group status asymmetry hypothesis. We believe we are the first to document that this asymmetry extends to generational status. To the extent that host adolescents are more assimilated than their immigrant counterparts, they may be more likely to believe that their ethnic group’s traits are compatible with the traits associated with “being American,” or they may be more likely to perceive themselves as having acquired more of the traits associated with being American the longer they are resident in the U.S., or both. The differences in these correlations were smaller at Time 2, reflecting the fact that the relationship between American identity and an ethnic identity component tended to increase for the immigrant groups, while remaining about the same for the host groups. Perhaps the intergroup contact experience increased immigrant adolescents’ perception of the compatibility of these two aspects of identity.

To our knowledge, our study is the first to distinguish ethnic identity exploration from ethnic identity affirmation/belonging when assessing the relationship of ethnic identity to American identity. Inclusion of both ethnic identity components allowed us to detect differences among the host groups. Specifically, for European-American host adolescents, American identity was significantly related to both aspects of their ethnic identity, in contrast to the two minority host groups (for which only one of the ethnic identity elements or the other was associated with American identity). For Asian-American host adolescents, the degree to which they identify as Americans may be compatible with the strength of their Asian identity because these two groups share a relatively high status in the school context, while at the same time, Asian cultures are not recognized as having made a strong contribution to the overall Southern California culture. The opposite pattern was obtained for Latino host adolescents. For them, the degree to which they identify as Americans was significantly related to their exploration of the traditions and customs of their Latino cultures, but unrelated to the sense of belonging to their ethnic group. Historically, the customs and traditions of Latinos in Southern California have had a relatively strong influence on the regional culture, so that learning about these contributions to the larger culture may increase Latinos’ identity as American, despite the fact that they may be lower in status than European Americans in the school context. The fact that different aspects of ethnic identity were related to American identity for these two ethno-generational groups shows the importance of disaggregating ethnic identity into its component parts when making statements about the compatibility of ethnic and national identity.

Research Advantages and Limitations

The compulsory nature of the course in which the intergroup contact occurred in the present study minimizes problems related to self-selection, but does not eliminate them, since participation in the surveys was contingent on written parental permission and student agreement. Nevertheless, participation rates averaged over 90%, which has positive implications for the generalizability of our findings. On the other hand, the classroom context presents a statistical/methodological challenge, because students are nested in classrooms. One solution to this issue is to do hierarchical linear modeling. This technique requires a large number of classes to be studied. As we accumulate more semesters of data, this analysis should be used.

A further limitation of the present study is that no subgroups were distinguished within either the “Asian-American” or “Latina/o” categories, despite the fact that each of these groups includes many different cultures, languages, and immigration histories. These factors could have influenced acculturation-related outcomes. A related limitation is that other important demographic and/or contextual variables (e.g., gender, SES, whether students live within or outside ethnic enclaves, etc.) were not taken into consideration. Including these could have provided a more complex understanding of the relationships among our study variables. Finally, the fact that there was a small number of European-American immigrant adolescents in our sample prevented us from comparing the host members of the dominant group with their immigrant counterparts.

Future Research

One of the implications of our work is that research on the “other side” of acculturation should not be limited to European-American host group members. Following legislation passed in 1965, the U.S. has received a relatively large number and high percentage of immigrants from Central and South America and Asian countries. Early waves of these post-1965 non-European immigrants are now an important part of the host culture to currently arriving groups, especially in cities of destination for the post-1965 immigrants. A challenge for acculturation researchers is to disentangle the experience of being a member of an ethnic minority in the U.S. from the experience of being an immigrant.

Furthermore, host ethnic minorities typically have higher social status than their immigrant counterparts and may feel a need to distinguish themselves from them. As such, they may be a source of intra-ethnic group prejudice and discrimination that has consequences for the acculturative experiences of new immigrants. This is a rarely-studied phenomenon. However, there is initial evidence (Saetermoe (2006) Unpublished paper), based on focus groups with 40 Latina/o second and third generation adolescents in Los Angeles, suggesting that members of this host generation tend to segregate themselves from immigrant Latina/o youth at their school and stigmatize them as “listening to rancheras and watching novellas.” In our research-in-progress, we are examining ethno-generational group membership as a source of intergenerational prejudice and perceptions of discrimination within racial/ethnic groups.

Conclusions

Using the concept of ethno-generational groups allowed us to clarify acculturation processes among adolescents who are born in the U.S. to U.S.-born parents, but whose heritage culture is non-European. We identified similarities and differences associated with the experience of being an Asian-American or Latino member of the host group, as compared to being a member of the Asian-American or Latino immigrant group and we demonstrated racial/ethnic group differences among the host group adolescents. These adolescent host group members of non-European heritage showed that their feelings about being an American were compatible with at least some aspects of their ethnic identity, in contrast to their immigrant group counterparts, for whom these two identities were unrelated.

However, the relationship between these identities for the Latino and Asian-American host adolescents was less comprehensive (i.e., it involved only one of two aspects of ethnic identity) than for their European-American counterparts. Their intermediate position may be indicative of a transitional role (e.g., en route to an American identity); may represent a fusion of identities (e.g., unconflicted, hyphenated American); or may connote an alternating form of biculturality (e.g., one that is compartmentalized with respect to the situation or with respect to the aspect of identity under consideration). On the one hand, these non-European host adolescents are likely to share visible physical features as well as a connection to their heritage culture with their immigrant group counterparts. On the other hand, they share a knowledge of the dominant culture with their European-American host group classmates. The extent to which each of these elements played a role in the pattern of results we obtained is unknown. It remains for future research, using the concept of ethno-generational group membership we have introduced in the present study, to assess the roles of these and other potential influences.