Who Am I? The Social Identities of Muslim-American Adolescents Madiha Tahseen and Charissa S.L. Cheah Abstract Muslim-American adolescents face unique developmental chal- lenges, due to the national and global negative spotlight on Islam, as they navigate identity formation in the mainstream American setting. However, the intersection of various social identities (SI) in the navigation of these challenges has been ignored in much so- cial-scientific literature. We examined the SI of Muslim-American adolescents using the multidimensional cluster analysis technique. Correlates of adolescents’ SI were also explored, specifically ado- lescents’ self-reported psychological well-being. One hundred and fifty Muslim adolescents were recruited from schools, mosques, and community organizations throughout Maryland. Follow-up analy- ses indicated that the High-Muslim/High-American adolescents reported the highest well-being and Moderate-Muslim/Undifferen- tiated-American cluster reported poorer well-being than the other clusters. Findings highlighted the importance of simultaneously assessing Muslim and American SIs, and the differential psycholog- ical benefits based on adolescents’ SI profiles. The article concludes by discussing implications for future research on Muslim adoles- cents’ successful adjustment. Despite the significant role of religion in development, the field of develop- mental psychology has largely ignored the impact of this key socio-cultural context on adolescents’ development.1 Muslim communities are also facing increased scrutiny due to the national and global spotlight on Islam. As such, Muslim-American adolescents face unique developmental challenges as they navigate identity formation in the mainstream American setting, Madiha Tahseen, PhD, is a Research Fellow at the Family and Youth Institute and a Post-Doctoral Researcher at the Culture, Child, and Adolescent Development Lab, at the University of Maryland, Baltimore County. Charissa Cheah, PhD, is a Profes- sor at the University of Maryland, Baltimore County, where she also serves as the Director of the Culture, Child, and Adolescent Development Lab. 32 The American Journal of Islamic Social Sciences 35:1 including lower social support, greater discrimination, and acculturative stress than other groups,2 all of which places them at a heightened risk for poor adjustment.3 Thus it is imperative to examine Muslim-American ad- olescents’ identity development and its association with their adjustment. The Muslim-American population is projected to double in size to about 8 million by 2030,4 with a majority (67 percent) being less than 40 years old.5 Despite these growing numbers, studies on Muslim-Americans are very limited. Most tend to be qualitative in nature and have focused on young adult populations, particularly the consequences of Muslim women’s veiling choices.6 Thus our understanding of how Muslim-American adoles- cents experience identity development in a heated sociopolitical context, and its impact on their outcomes, is limited. To address these limitations, the present study had two aims. The first aim was to adopt a multidimensional approach to examining identi- ty among Muslim adolescents by using cluster analysis to create profiles of their social identities. We also compared the resulting identity clusters on adolescents’ psychological well-being. Our second aim was to provide professionals, communities, and policymakers with recommendations for promoting the positive identity development of Muslim-American youth. Intersecting Social Identities The Muslim community at large can be defined as a diaspora, or a geo- graphically dispersed group which is unified by a common identity,7 as well as an ‘Ummah’, a transnational community of Muslims in which members bear a collective representation.8 These aspects result in Muslim adolescents’ heightened sense of membership with the Muslim community,9 which be- comes more salient through their interactions in various contexts within the Muslim diaspora. Meanwhile, due to Islamophobia and the treatment of Muslims as “others” in legal policies, media dialogue, and educational institutions,10 Muslim identity has been referred to as “identity formation under siege”. Negative social imagery within various contexts (e.g. media and classrooms) results in greater pressure to negotiate their identification with the “Muslim” and “American” group labels than other racial/ethnic groups, and impacts Muslim adolescents’ self-image.11 Intersecting Identities The compatibility between Muslim adolescents’ American and Muslim social identities remains unclear. In contrast to the “civilization clash” ap- 33 proach,12 which posits that individuals are unable to reconcile two different cultures, an intersectional perspective considers identity as encompassing the confluence of multiple social identities in each individual, which are interdependent and time- and context-dependent.13 Similarly, the hyphen- ated identities framework suggests that Muslim-American adolescents ne- gotiate multiple allegiances to and experience multiple stresses from their Muslim communities and the mainstream US society.14 Indeed, researchers have demonstrated that Muslim adolescents work actively towards creating hyphenated or bicultural identities that are uniquely comprised of Muslim and mainstream components.15 Although preliminary findings suggest that Muslim adolescents can endorse a strong sense of national belonging and embrace a non-Muslim lifestyle,16 we still have a very limited understand- ing of Muslim adolescents’ identification with the larger society (i.e. nation- al identity)17 due to the limited research on this type of identity. Moreover, past research has used the cut-off score method to create identity profiles, which is problematic because adolescents are then cate- gorized based on a single score representing their identity profile, and the data are split into identity groups in samples where such heterogeneity may not exist.18 Thus, identity is treated as a dichotomous variable and sample variance is ignored, potentially mislabeling the identity patterns that may exist among Muslim adolescents. Finally, ethnic versus religious compo- nents of identity have often been confounded in past studies, which limits our understanding of Muslim identity as such, and its generalizability to Muslim adolescents’ religious identity experiences across ethnic and na- tional boundaries. To address these limitations, we adopted an intersectional framework which focuses on how adolescents experience their belonging in various so- cial categories.19 Consistent with this framework, we used a person-centered approach to identity through cluster-analytic statistics,20 which creates clus- ters of individuals based on the pattern of their responses to several differ- ent items across a variety of variables.21 Instead of dichotomizing identity by creating groups based on cut-off scores, the cluster analysis method treats identity as a continuous variable and provides insight about the variance that exists in the sample. We used this technique to examine adolescents’ pattern of responses across the two identity scales, thereby allowing the simultaneous assessment of both components of social identity (i.e. Muslim and American cultures). The resulting clusters served as the identity profiles used in subse- quent analyses. Due to the lack of existing research on identity patterns of Muslim adolescents, our analyses were exploratory in this regard. Tahseen and Cheah: Who Am I? 34 The American Journal of Islamic Social Sciences 35:1 Psychological Well-being Extensive research on minority adolescents demonstrates the pro- tective effects of identifying with one’s group in response to perceptions of mistreatment.22 The rejection-identification model posits that group members whose identities are contested by others will value their in-group membership more than their out-group membership,23 which will result in protective effects. Strengthened group identification with one’s own group provides individuals with social support, a sense of belonging with ingroup members, and coping resources.24 Indeed, researchers have found that religious identification was related to Muslim adolescents’ positive outcomes, and negatively related to nega- tive outcomes.25 However, the majority of past researchers have primarily used religious affiliation as indicative of group identification. Past research has also largely ignored the influence of Muslim adolescents’ national identity in their well-being. This is concerning, given that self-complexity perspectives propose that multiple self-representations and group iden- tities can provide beneficial, protective effects in the face of stereotyping, prejudice, and discrimination.26 Indeed, research among ethnic minorities supports this finding,27 however these associations have not been explored among Muslim youth. The influence of simultaneous identification with multiple groups on well-being among Muslim adolescents too has not yet been examined. Thus, we sought to understand how adolescents’ Muslim and American identities intersect to influence their well-being. We expect- ed that Muslim adolescents with stronger Muslim and American collective identities would report greater well-being. Method Participants One-hundred and fifty Muslim adolescents (M = 15.73 years, SD = 1.79 years; 11.49 to 19.03 years) participated in the present study (Table 1). The sample comprised fifty-two males (35%) and ninety-eight females (65%). Adolescents were primarily from South-Asian (e.g. Pakistan, India, Ban- gladesh; 52%) or Arab (e.g. Palestine, Egypt, Morocco; 17%) ethnic back- grounds, which is representative of the ethnic makeup of the communities in the Baltimore-Washington metropolitan region. Adolescents were also from East-Asian (7%), African/Black (5%), African-American (7%), Bi-ra- cial (5%) and Other (6%) backgrounds. All of the 20% of adolescents who were first-generation had migrated to the United States at less than six years 35 of age. The majority of adolescents were second-generation immigrants (69%), whereas 2% were third-generation and 9% were fifth-generation im- migrants. The majority (61%) of adolescents attended public school, where- as 20% attended private school, 14% were in college, and 5% were homes- chooled. Post-hoc analyses revealed no significant differences between the sub-groups of adolescents based on school-type across all study variables. Almost all the adolescents lived in two-parent households (95%), in which both parents were married (91%), and contained four children or less (79%). Adolescents reported high levels of education for fathers, such that 38% of fathers had a standard undergraduate degree, and 33% had a graduate or professional degree. Although mothers had similar levels of un- dergraduate education (33%), fewer mothers had a graduate degree (14%) and 25% had completed partial college. Both parents have lived in the US for an average of 22 years (SD = 11.90). In order to understand adolescents’ engagement and interaction with the Muslim community, adolescents were asked to report on their attendance of Muslim classes, religious services, and activities held by Muslim organiza- tions. Almost half of the adolescent sample (47%) attended the mosque (Ar- abic masjid) for prayers and religious services once or twice a week. About a third of the sample (39%) attended the masjid multiple times a week, whereas 14% never or rarely attended the masjid for prayers or services. Similarly, approximately 57% of adolescents participated in activities held by Muslim organizations once or twice a week, whereas about 20% attended these ac- tivities multiple times a week, and 22% never attended at all. Finally, 49% of adolescents attended Muslim classes once or twice a week, whereas the ma- jority (51%) did not attend any Muslim classes. Sub-samples of adolescents were created within the various levels of engagement with Muslim commu- nity variables (i.e. frequently attending masjid versus infrequent attendance), type of schooling (e.g. private versus public) and generational status (i.e. second-generation versus first-generation) separately. These subsamples (e.g. students who attend public school versus private school) were then compared on demographic, identity, and respective correlates (i.e. psychological func- tioning). Post-hoc t-tests revealed no significant differences between the sub- samples of adolescents across all study variables. Procedures Participants were recruited using convenience and snowball sampling methods in person and online, from Muslim organizations, public and private Tahseen and Cheah: Who Am I? 36 The American Journal of Islamic Social Sciences 35:1 schools, and community centers in the Baltimore-Washington Metropolitan area. After obtaining parental consent, adolescents were directed to complete the surveys on a secure and confidential online website. Participants were provided with the researcher’s contact information in case they had any ques- tions regarding the questionnaires. Upon completion of the questionnaires, adolescents were reimbursed $20 and received community service hours. Measures All of the variables that were used in the present research were assessed through online questionnaires and administered in the English language. Social Identity Adolescents reported on their affective attachment to their social iden- tity and the importance of their social identity. The Multigroup Ethnic Iden- tity Measure-Revised28 (MEIM-R) was used to assess adolescents’ affective attachment (3 items; “I feel a strong attachment towards my Muslim (or American) cultural group”). Adolescents rated the items using a 1 (strong- ly disagree) to 5 (strongly agree) Likert scale. The Collective Self-Esteem Scale-Revised29 (CSES-R) was used to assess adolescents’ perception of the explicit importance of their Muslim and American group membership to their self-concept (4 items; “Being a Muslim is an important part of who I am”). Adolescents rated each item using a 1 (never) to 5 (always) Likert scale. For all of the measures of social identity, adolescents were instructed to think of their Muslim and American group membership separately. High scores on each measure reflected greater endorsement of overall American and Muslim identity. The CSES-R and MEIM-R demonstrated adequate re- liability in past research.30 In the present sample, α = .84 for both American and Muslim social identity. Psychological Well-being The Psychological Well-Being Scale was administered to assess ado- lescents’ psychological well-being (PWBS).31 This measure captures six dimensions of well-being: autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance using a scale of “1” (Strongly Disagree) to “7” (Strongly Agree). An overall psychological well-being score was created by summing adolescents’ scores on all of the subscales. The PWBS has demonstrated adequate internal con- sistency across multiple diverse samples, with alphas ranging from .77 to .94.32 For the present sample, α = .76. 37 Results Aim One: Identifying Identity Clusters The hierarchical clustering was used to explore the clusters that exist in the data, because it does not require a preset number of clusters and allows the number of clusters to vary based on the data, resulting in a cluster solu- tion that is the best fit to the data.33 Ward’s 1963 method of clustering opti- mizes the minimum amount of variance within clusters, by combining those entities that have the smallest distance between them.34 The following identity variables were submitted to the Ward’s method: (1) American social identity and (2) Muslim social identity. The identity measures were standardized prior to conducting the analysis. Next, three strategies were applied to determine the optimal number of clusters that should be retained: (1) interpretation of the dendrogram35; (2) application of Mojena’s Rule One36; and (3) an exam- ination of the fusion coefficients.37 From these three strategies, the six-cluster solution was deemed the most optimal fit to the data and was retained as the grouping variable for all subsequent analyses.38 Clusters 5 and 6 contained eight and seven participants respectively. However, due to the limited amount of research on Muslim-American ado- lescents, and given that retaining cluster 5 and 6 for this specific set of anal- yses aids in our understanding of the relative makeup of each cluster, these clusters were retained for follow-up analyses in Aim 1 (i.e. labeling of each cluster). However, clusters 5 and 6 were not included in the remaining anal- yses involving psychological well-being (Aim 2). The findings presented below involving clusters 5 and 6 should be interpreted in light of this issue. Consistent with past research, two strategies were used to describe the clusters.39 First, two one-way between-group (BG) ANOVAs were conducted to examine mean differences between the clusters on each of the identity fac- tors. Second, independent samples t-tests were conducted to examine wheth- er the clusters’ means differed from the sample mean within each identity fac- tor. Univariate results are presented in the description of each of the clusters below, and in Table 2 and Figure 1. Using these mean comparison analyses, the clusters were described using the High, Moderate, Low, or Undifferentiat- ed labels. The difference between the labels of ‘Moderate’ and ‘Undifferentiat- ed’ should be noted. Adolescents in the Moderate clusters scored higher than the sample mean and/or differed from the other clusters in a clearer pattern than Undifferentiated adolescents. Further, in absolute terms, the means for the Moderate clusters (e.g., 3.72, 3.67) were higher on the identity dimensions than the means for the Undifferentiated clusters (e.g., 2.92, 2.66). Tahseen and Cheah: Who Am I? 38 The American Journal of Islamic Social Sciences 35:1 Cluster 1 (N =38; 25% of the sample). Cluster 1 was the second largest cluster in the sample. Adolescents in the first cluster scored higher on Muslim social identity than adolescents in Clusters 2, 4, and 6, and the sample mean on this factor, but not Clusters 3 and 5. Regarding American social identity, Cluster 1 adolescents scored lower than adolescents in Clusters 2 and 3, and the sample mean. However, these adolescents scored higher than adolescents in Cluster 5 on American identity, but did not differ from adolescents in Clus- ter 4 or Cluster 6. Thus, with relatively high scores on Muslim social identity and an unspecified pattern of responses on American identity, this cluster was labeled “High-Muslim and Undifferentiated-American Social Identity.” Cluster 2 (N =28; 19% of the sample). Cluster 2 was smaller than Clusters 1 and 3. Adolescents in Cluster 2 scored lower on Muslim social identity than adolescents in Clusters 1, 3, 5, and the sample mean. However, Cluster 2 adolescents scored higher than Cluster 6 but did not differ from Cluster 4 adolescents on Muslim social identity. With respect to American social identity, Cluster 2 adolescents scored higher than those in Clusters 1, 4, 5, 6, and the sample mean, but did not differ from adolescents in Cluster 3. Thus, with moderately low scores on Muslim identity, and relatively high scores on American identity, this cluster was labeled “Moderate-Muslim and High-American Social Identity.” Cluster 3 (N =53; 35% of the sample). Cluster 3 was the largest clus- ter found in the sample. Adolescents in this cluster scored high on both Muslim and American identity factors. With respect to Muslim social iden- tity, Cluster 3 adolescents scored higher than adolescents in Clusters 2, 4, 6, and the sample mean, but did not differ from adolescents in Cluster 1 and Cluster 5. Regarding American social identity, adolescents in Cluster 3 scored higher than adolescents in Cluster 1, 4, 5, 6, and the sample mean, but did not differ from adolescents in Cluster 2. Thus, with high scores on both Muslim and American identity factors, this cluster was labeled “High Muslim and American Social Identity.” Cluster 4 (N =16; 11% of the sample). Adolescents in Cluster 4 scored relatively low on both Muslim and American identity factors. Regarding Muslim social identity, Cluster 4 adolescents scored lower than adolescents in Clusters 1, 3, 5, and the sample mean. However, these adolescents scored higher than Cluster 6 but did not differ from Cluster 2 adolescents on this factor. With respect to American social identity, Cluster 4 adolescents scored lower than Clusters 2, 3, and the mean, but higher than Cluster 5. Cluster 4 adolescents did not differ from Cluster 1 and Cluster 6 adolescents on Amer- ican identity. Thus, with moderately low scores on the Muslim identity fac- 39 tor, and an unclear pattern on the American identity factor, this cluster was labeled “Moderate-Muslim and Undifferentiated-American Social Identity.” Cluster 5 (N =8; 5% of the sample). Adolescents in Cluster 5 was the second smallest cluster in the sample. Cluster 5 adolescents scored higher on Muslim social identity than adolescents in Clusters 2, 4, and 6, and the sample mean, but did not differ from Clusters 1 and 3. Regarding American social identity, Cluster 5 adolescents scored lower than all of the clusters and the sample mean. Thus, with high scores on Muslim identity and low scores on American identity, this cluster was labeled “High-Muslim and Low-American Social Identity.” Cluster 6 (N =7; 5% of the sample). Cluster 6 was the smallest cluster in the sample. Adolescents in Cluster 6 scored lower than all of the other clusters and the sample mean on Muslim social identity. Regarding American social identity, Cluster 6 adolescents scored lower than adolescents in Cluster 2 and 3, but higher than Cluster 5 adolescents, similar to Cluster 1 adolescents. However, Cluster 6 adolescents did not differ from the sample mean, or from adolescents in Clusters 1 or 4. Thus, with low scores on Muslim identification and an unclear pattern of scores on American identity, this cluster was labeled “Low-Muslim and Undifferentiated-American Social Identity”. Aim Two: Psychological Well-being of the Identity Clusters The second aim of the present study was to examine the psychological well-being associated with the resulting social identity clusters. As stated above, Clusters 5 and 6 were excluded from this set of analyses. Psychological Well-Being. A BG-ANOVA revealed that the iden- tity clusters differed on self-reported psychological well-being (Table 3). Cluster 4 (Moderate-Muslim/Undifferentiated-American) adolescents re- ported lower levels of psychological well-being than Cluster 1 adolescents (High-Muslim/Undifferentiated-American) and Cluster 3 adolescents (High-Muslim/High-American). There were no significant differences be- tween Cluster 2 (Moderate-Muslim/High-American) and the other clusters on psychological well-being. Discussion The present study examined the complex social identity experiences of Muslim-American adolescents using a cluster analysis statistical approach, as well as the well-being associated with each identity profile. Tahseen and Cheah: Who Am I? 40 The American Journal of Islamic Social Sciences 35:1 General Patterns of Social Identity Results from the cluster analysis technique indicated the presence of six social identity profiles: (1) High-Muslim and Undifferentiated-Amer- ican (n = 38), (2) Moderate-Muslim and High-American (n = 28), (3) High Muslim and High-American (n = 53), (4) Moderate-Muslim and Undifferentiated-American (n = 16), (5) High-Muslim and Low-Amer- ican (n = 8), (6) Low-Muslim and Undifferentiated-American (n =7). In general, the findings from the cluster analysis method demonstrated that Muslim-American adolescents do indeed live in the middle of the hyphen, where they navigate the intersection of their Muslim and American iden- tities. Consistent with past research, Muslim-American adolescents exhib- ited high levels of Muslim identity (66%) in the present sample.40 Thus, in line with the rejection-identification model,41 despite having their identities questioned in a heated socio-political context, Muslim-American adoles- cents are still able to identify strongly with the Muslim collective. Interestingly, Muslim adolescents’ American identity levels varied across all of the High Muslim identity profiles (i.e. High-Muslim/Undif- ferentiated-American, High-Muslim/High-American, and High-Muslim/ Low-American). These distinct patterns of American identity demonstrate the heterogeneity that exists among Muslim-American adolescents as they explore the intersection of their Muslim and American identities. For some High-Muslim adolescents, their identity negotiations are not solely focused on alienation but also include desires to engage and attach with mainstream Americans (i.e. High-Muslim/High-American). Other High-Muslim ado- lescents are still exploring their American identities and have not yet turned away from the mainstream culture (i.e. High-Muslim/Undifferentiat- ed-American). Contrarily, some High-Muslim adolescents dissociate with American culture and emphasize their Muslim group membership in order to achieve a sense of belonging (i.e. High-Muslim/Low-American). A significant percentage of Muslim-American adolescents (41% across three profiles) exhibited an Undifferentiated-American identity, which sug- gests that a large group of Muslim-American adolescents have not committed to a clear sense of American identity. The present sample comprised Muslim students enrolled in late middle school and high school, who are facing var- ious elements of American culture which conflict with Muslim values and practices (e.g., dating, experimenting with alcohol and drugs) during their daily interactions.42 Thus, they may be in the process of exploring how these conflicting aspects of American culture fit with their American identities. 41 Specific Identity Profiles The High-Muslim/High-American profile was one of the most prevalent profiles (35%) in the present sample. Similar to the bicultural profile found among Muslim adolescents across the world,43 High-Muslim/High-Ameri- can adolescents indeed ascribe great importance to both cultural groups and reported a strong sense of attachment to both groups. This finding is in stark contrast to the clash hypothesis, or the idea that due to the significantly neg- ative attention Islam has received in the sociopolitical context, Muslim ado- lescents are unable to maintain connections with both American and Muslim groups.44 Contrariwise, adolescents in the High-Muslim/High-American profile displayed hyphenated selves, or an amalgamation of Muslim and American components, supporting the idea that Muslim-American adoles- cents indeed ascribe great importance to both cultural groups. Also, high lev- els of simultaneous identification with both groups suggests that the potential intersectionality between American and Muslim culture does not serve as a hindrance to adolescents’ attachment with either cultural group. Moreover, the presence of both High-Muslim/High-American and Moderate-Muslim/High-American profiles demonstrates that Muslim adolescents hold relatively strong American and Muslim identities in dif- ferent ways. This finding is consistent with recent research, which uncov- ered two types of bicultural identities among Arab-American adolescents: Moderate Bicultural and High Bicultural identity.45 The Muslim identity of Moderate-Muslim/High-American adolescents may be at a different devel- opmental stage than the Muslim identity of High-Muslim/High-American adolescents. These adolescents may be exploring and resolving the mixed messages they experience regarding their membership in the Muslim group. Moderate-Muslim/Undifferentiated-American adolescents do not have a strong emotional bond with either Muslim or American identity, and rate each group as moderately important for their overall identity. Diffuse adolescents are described as lacking a clear orientation and being confused about their identifications,46 being uncommitted to a specific purpose in their lives and feeling socially isolated.47 Thus, similar to the diffuse profile, the Moderate-Muslim/Undifferentiated-American profile may comprise adolescents who have not yet committed to either Muslim or American group. Also, the inconsistent pattern of identification may result in a hybrid form of identity that is not reflected in either Muslim or American identity separately, which has been found in past research among second-genera- tion Muslims.48 Tahseen and Cheah: Who Am I? 42 The American Journal of Islamic Social Sciences 35:1 The final profile, Low-Muslim/Undifferentiated-American, was the smallest cluster (5%) in the sample. Low-Muslim/Undifferentiated-Ameri- can adolescents seem to prefer an identity profile that falls between the “na- tional” and “marginalized” profile in identity research.49 These adolescents did not score sufficiently low on both American and Muslim components to be characterized as marginalized. Although their American identity was higher than their Muslim identity, the difference was not as large as the typical pattern found in “national” profiles. Low-Muslim/Undifferen- tiated-American adolescents may be exploring their American identities; however, they have not achieved a sense of clarity regarding the American group. The low proportion of adolescents in this profile is consistent with the majority of past research which shows that Muslim adolescents do not frequently solely identify with the national group. Well-being associated with the Social Identity Profiles. Adolescents who were high on Muslim identity (i.e. High-Muslim/Undifferentiat- ed-American and High-Muslim/High-American profiles) reported greater well-being than adolescents who were low on both Muslim and American identities (Moderate-Muslim/Undifferentiated-American profile). In line with our hypotheses and past research, a strong attachment to the Muslim culture had positive effects on adolescents’ psychological functioning.50 Con- sistent with the rejection-identification paradigm, Muslim adolescents whose identities are contested by others still valued their in-group membership (i.e. Muslim), which was related to their positive well-being. Strengthened group identification with one’s own group and a shared sense of collective identity may provide adolescents with social support, sense of belonging with in- group members, and coping resources, which may protect adolescents from the negative effects of growing up in a heated sociopolitical context.51 The greater well-being of adolescents in the High-Muslim/High-Amer- ican profile is consistent with the research on bicultural profiles found among Muslim adolescents in other studies.52 In line with self-complexity perspectives, adolescents in this profile have high attachment to multiple identities and are able to successfully navigate a variety of social settings. They have amassed high identity capital, which refers to the tangible (e.g. education) and intangible (e.g. psychological capacity, self-evaluation) re- sources that adolescents have acquired through various identity interac- tions in their daily lives.53 Identity capital allows High-Muslim/High-Amer- ican adolescents to express their various identities in ways that helps them achieve material or emotional goals,54 and serves as a resource in the face of religious-related identity threats. Thus, the High-Muslim/High-American 43 identity can be empowering for Muslim-American adolescents, resulting in greater well-being. Consistent with past research, the lower level of psychological well-being reported by Moderate-Muslim/Undifferentiated-American adolescents sug- gests that adolescents who exhibit diffuse identity styles are at risk for poor outcomes.55 These adolescents may be engaging in the broad type of explora- tion, in which adolescents consider various identity alternatives without any definite commitments, rather than exploration in depth in which adolescents are intently focused on a previously formed commitment.56 They may feel socially isolated, and given that this type of exploration is associated with depressive symptoms in past research,57 Moderate-Muslim/Undifferentiat- ed-American adolescents may be experiencing an identity crisis, resulting in lower psychological well-being. This finding further supports past findings that youth with diffuse identity styles are at risk for poor outcomes.58 Limitations and Future Directions Several limitations of the present research should be noted. The small sample sizes of Clusters 5 and 6 may have affected the comparisons between the identity profiles. Second, due to the non-random sampling recruitment procedures used and the sample’s level of Muslim identity, generalizabili- ty to other subgroups of Muslim adolescents was limited. Future research should use various sampling methods to increase the heterogeneity and representativeness of the sample. Third, our data were self-reported using the cross-sectional design. Al- though adolescents actively endorse various identity profiles which affect their well-being,59 it may be that their psychological functioning predicts their endorsement of various identity profiles. Future research should adopt a longitudinal design to assess identity progression and the temporal prece- dence of identity over psychological well-being. The final set of limitations pertains to our interpretation of the six clus- ters. First, consistent with previous research60 and due to a small sample size, a within-sample approach was adopted when describing the clusters, such that adolescents in the clusters were compared to other adolescents in the clusters, as well to the overall sample means on the identity measures. For example, Moderate-Muslim/High-American adolescents’ American component was labeled as High and their Muslim component was labeled as Moderate, although there was a relatively small difference between these two components within this cluster. The overall sample was relatively lower on the American component than the Muslim component. Moderate-Mus- Tahseen and Cheah: Who Am I? 44 The American Journal of Islamic Social Sciences 35:1 lim/High-American adolescents’ American identity score was higher than the other clusters and the sample; however, it was less than a standard de- viation away from the overall sample mean on this variable (Table 2). Thus, their American identity may not be that high in absolute terms or com- pared to adolescents in previous research.61 Finally, we tested the clusters on the same variables used to create the clusters, which may have resulted in highly significant differences between the clusters. Although acceptable, future research should validate the retained cluster solution against an external set of variables that represent Muslim and American social identity, and replicate the analysis in another sample.62 Recommendations Findings from the present research have implications for mental health service providers, schools, community organizations, and policymakers at various levels. Mental Health Providers. First, in line with the Code of Ethics for psychologists and for social workers, mental health service providers must obtain relevant knowledge and training to maintain competence in work- ing with clients of diverse backgrounds.63 Findings regarding the identity profiles of Muslim adolescents can educate mental health professionals about the unique situation of combined Muslim and American identities so that they are able to recognize and identify these profiles in their clients, as well as identify Muslim adolescents at risk for poor outcomes, such as the Moderate-Muslim/Undifferentiated-American profile. Given the finding of high Muslim identity among adolescents in the present sample, clinicians need to increase their knowledge with the teachings of Islam and consider the role of spiritual healing for Muslim adolescents.64 Schools. The current finding of various levels of moderate to high iden- tification with both American and Muslim identities also has implications for the schooling context. First, schools should engage in programs that facilitate Muslim students’ identifications with both Muslim and American groups. For example, Muslim private schools should offer and promote their students’ interactions with mainstream American environments, perhaps through the use of extracurricular activities and through religious curricula which contex- tualize Islam within contemporary American society.65 Public schools should create inclusive environments that support multiculturalism and diversity. This can be achieved in the following ways: (1) greater physical displays of diversity, such as posters on the walls, books in the library, and including the history of Muslims in America within curriculums; (2) provide staff with religious and 45 cultural competency trainings to support Muslim-American students, and (3) offer greater support for students’ participation in Muslim Student Associations and religious accommodation, such as space for offering prayers, excused ab- sences for religious holidays. More importantly, schools should develop and offer anti-bullying messaging which focus on increasing awareness of Islam and Muslims among their staff and general student body, in order to counter the civilization-clash concept that is promoted in the media.66 Communities. At the community level, findings regarding the collec- tive identity style endorsed by adolescents can be disseminated to organi- zations and communities that serve Muslim youth. These agencies should promote identity development among their Muslim youth, by offering greater support for youth groups and by creating seminars or programs that provide opportunities and skills that Muslim adolescents need to navigate both Muslim and American settings. Finally, and perhaps most important- ly, given that moderate to high American identities were endorsed by our sample of Muslim adolescents, Muslim communities should adopt positive attitudes towards the mainstream American culture and model positive in- teractions with American society to help Muslim adolescents integrate their American identities with their religious identities. Policymakers. Finally, the findings from the present study also have implications for policymakers, governmental organizations and think- tanks, such as the Office of Minority Health, 4-H Youth Development Program, Child Trends, and Institute for Social Policy and Understanding. These entities should use the current findings of the healthy psychological functioning of adolescents with high American and Muslim identities to in- crease funding of comprehensive research programs on Muslim adolescent development, as well funds to develop and evaluate programs that promote adolescents’ involvement in both Muslim and American settings, which fosters their psychological well-being. Despite these limitations, the present research served as a multifaceted examination of the social identities of Muslim-American youth residing in a politically stressed context. The person-centered approach to identity provided an exploration of the complex nature of group identity among minority Muslim adolescents. The difference in psychological well-being among the identity clusters furthers our understanding of the functionality of various identity profiles. Future research can build on this study’s find- ings through various research methods in order to better understand the risk and protective factors associated with Muslim adolescents’ well-being in the United States. Tahseen and Cheah: Who Am I? 46 The American Journal of Islamic Social Sciences 35:1 Table 1. Demographic Characteristics of the Sample (Adolescent Report) Variable Category Percentage Education Status 8th grade 24 (16%) High school 104 (70%) First-semester freshmen 22 (14%) Type of School Public secondary school 91 (61%) Private secondary school 29 (19%) Home-Schooled 8 (5%) College 22 (14%) Ethnicity South-Asian 78 (52%) Arab 25 (17%) East-Asian 11 (7%) African/Black 7 (5%) African-American 10 (7%) Bi-racial 8 (5%) Other 9 (6%) Generational Status First-generation 30 (20%) Second-generation 103 (69%) Third-generation 3 (2%) Fourth-generation 1 (1%) Fifth-generation 13 (9%) Paternal Education Less than middle school 6 (4%) High school graduate 14 (9%) Partial college 15 (10%) Undergraduate degree 57 (38%)   Graduate/Professional degree 49 (33%) Maternal Education Less than middle school 9 (6%) High school graduate 25 (17%) Partial college 37 (25%) Undergraduate degree 50 (33%) Graduate/Professional degree 21 (14%) 47 Table 2. Mean Differences between Clusters on Muslim and American Social Identity Muslim Social Identity American Social Iden- tity F (5, 144) = 60.25***, ηp 2 = .76 F (5, 144) = 83.35***, ηp 2 = .73 M (SD) M (SD) High-M/Undifferentiated-A (Cluster 1; N = 38) 4.58 (.22) 2.92 (.37) Sample Mean 4.24 (.67) 3.37 (.77) t = 9.42***, d = .68 t= -7.54***, d = .74 Moderate-M/High-A (Cluster 2; N = 28) 3.72 (.35) 3.73 (.52) Sample Mean 4.24 (.67) 3.37 (.77) t= -7.76***, d = .97 t= 3.70**, d = .55 High-M/High-A (Cluster 3; N = 53) 4.60 (.34) 4.02 (.30) Sample Mean 4.24 (.67) 3.37 (.77) t= 7.83***, d = .68 t= 15.49***, d = 1.11 Moderate-M/Undifferentiated-A (Cluster 4; N = 16) 3.76 (.23) 2.66 (.43) Sample Mean 4.24 (.67) 3.37 (.77) t= -8.43***, d = .96 t= -6.60***, d = 1.14 High-M/Low-A (Cluster 5; N = 8) 4.69 (.29) 1.71 (.35) Sample Mean 4.24 (.67) 3.37 (.77) t= 4.40*, d = .96 t= -13.29***, d = 2.78 Low-M/Undifferentiated-A (Cluster 6; N = 7) 2.31 (.75) 2.90 (.55) Sample Mean 4.24 (.67) 3.37 (.77)   t= -6.78**, d = 2.71 t= -2.23, ns, d = .70 Note. “M” refers to Muslim identity, and “A” refers to American identity. *** p < .001; ** p < .01; * p < .05 Tahseen and Cheah: Who Am I? 48 The American Journal of Islamic Social Sciences 35:1 Table 3. Well-being F (3, 129) = 3.60, p < .05, ηp 2 = .08 Cluster M (SD) High-M/Undifferentiated-A (N = 38) 5.18 (.67) Moderate-M/High-A (N = 29) 5.03 (0.60) High-M/High-A (N = 53) 5.17 (0.72) Moderate-M/Undifferentiated-A (N = 16) 4.60 (0.38) High-M/Low-A (N = 8) — Low-M/Undifferentiated-A (N = 7) — Cluster differences in adolescents’ report of psychological well-being Note. “M” refers to Muslim identity, and “A” refers to American identity. 49 Figure 1. Six cluster identity solution based on z-scores Note. In the X-axis labels, the abbreviation “U-” refers to the Undifferentiated style of identity. Tahseen and Cheah: Who Am I? 50 The American Journal of Islamic Social Sciences 35:1 Endnotes 1. Sadia R. Chaudhury and Lisa Miller, “Religious Identity Formation Among Bangladeshi American Muslim Adolescents,” Journal of Adolescent Research 23, no. 4 (2008): 383-410. 2. Sameera Ahmed and Maha Ezzeddine, “Challenges and Opportunities Facing American Muslim Youth,” Journal of Muslim Mental Health 4, no. 2 (2009): 159-74. 3. Selcuk R. Sirin and Michelle Fine, Muslim-American Youth: Understanding Hyphenated Identities through Multiple Methods (New York: New York Uni- versity Press, 2008) 4. Pew Research Center, “Muslim American: No Signs of Growth in Alienation or Support for Extremism” (2011), http://www.people-press.org/files/lega- cy-pdf/Muslim%20American%20Report%2010-02-12%20fix.pdf. 5. Selcuk R. Sirin and Aida B. Balsano, “Editors’ Introduction: Pathways to Identity and Positive Development Among Muslim Youth in the West,” Ap- plied Developmental Science 11, no. 3 (2007): 109-11. 6. Sirin and Fine, Muslim-American Youth. 7. Riva Kastoryano, “Integration and Collective Identities of Immigrants in France and Germany,” Journal of Ethnic Studies 19, no. 3 (1998): 51–64. 8. Cheryl S. Al-Mateen and Aneeta Afzal, “The Muslim Child, Adolescent, and Family,” Child and Adolescent Psychiatric Clinics of North America 13, no. 1 (2004): 183-200. 9. Frederick Denny, Islam and the Muslim Community (New York: HarperCol- lins, 1993). 10. Amber Haque, “Islamophobia in North America: Confronting the Menace,” in Confronting Islamophobia in Educational Practice, ed. B. van Driel (Stoke- on-Trent, UK: Trentham Books, 2004): 1-18. 11. Carola Suárez-Orozco, “Formulating Identity in a Globalized World,” in Globalization: Culture and Education in the New Millennium, ed. M.C. Suárez-Orozco and D.B. Qin-Hilliard (Berkeley, CA: University of California Press, 2004), 173-202. 12. Samuel P. Huntington, Who Are We: The Challenges to America’s National Identity (New York, NY: Simon & Schuster, 2004). 13. Dee Watts-Jones, “Location of Self: Opening the Door to Dialogue on Inter- sectionality in the Therapy Process,” Family Process 49, no. 3 (2010): 405-420. 14. Sirin and Fine, Muslim-American Youth. 15. Chaudhury and Miller, Religious Identity Formation; Susan L. Ketner, Marjo J. Buitelaar, & Harke A. Bosma, “Identity Strategies among Adolescent Girls of Moroccan Descent in the Netherlands,” Identity: An International Journal of Theory and Research 4 (2004): 145-169; Sirin and Fine, Muslim-American Youth. 51 16. Sameera Ahmed and Linda A. Reddy, “Understanding the Mental Health Needs of American Muslims: Recommendations and Considerations for Practice,” Journal of Multicultural Counseling and Development 35, no. 4 (2007): 207-18; Karen Phalet and Derya Güngör, “Moslim in Nederland: Religieuze dimensies, etnische relaties en burgerschap” (The Hague, Nether- lands: Social and Cultural Planning Office, 2004). 17. Maykel Verkuyten, Jochem Thijs, and Gonneke Stevens, “Multiple Identities and Religious Transmission: A Study Among Moroccan-Dutch Muslim Ad- olescents and Their Parents,” Child Development 83, no. 5 (2012): 1577-590. 18. Judith Arends-Tóth and Fons J.R. Van de Vijver, “Conceptualization and As- sessment of Acculturation,” in Acculturation and Parent-Child Relationships: Measurement and Development, ed. M. H. Bornstein and L. R. Cotè (Mah- wah, NJ: Erlbaum, 2006), 33-62. 19. Nira Yuval-Davis, The Politics of Belonging: Intersectional Contestations (Lon- don, UK: SAGE Publications, 2011). 20. Mark S. Aldenderfer and Roger K. Blashfield, Cluster Analysis (Newbury Park: Sage, 1984). 21. Lars R. Bergman and David Magnusson, “A Person-Oriented Approach in Research on Developmental Psychopathology,” Development and Psychopa- thology 9, no. 2 (1997). 22. James L. Furrow, Pamela Ebstyne, and Krystal White, “Religion and Positive Youth Development: Identity, Meaning and Prosocial Concerns,” Applied De- velopmental Science 8, no. 1 (2004): 17-26. 23. Nyla R. Branscombe, Michael T. Schmitt, and Richard D. Harvey, “Perceiving Pervasive Discrimination among African Americans: Implications for Group Identification and Well-being,” Journal of Personality and Social Psychology 77, no. 1 (1999): 135-49. 24. Nazilla Khanlou, Jane G. Koh, and Catriona Mill, “Cultural Identity and Ex- periences of Prejudice and Discrimination of Afghan and Iranian Immigrant Youth,” International Journal of Mental Health and Addiction 6, no. 4 (2008): 494-513. 25. Hisham M. Abu-Rayya, and Maram Hussien Abu-Rayya, “Acculturation, Re- ligious Identity, and Psychological Well-being among Palestinians in Israel,” International Journal of Intercultural Relations 33, no. 4 (2009): 325-31. 26. Margaret Shih and Diana T. Sanchez, “Perspectives and Research on the Pos- itive and Negative Implications of Having Multiple Racial Identities,” Psycho- logical Bulletin 131 (2005): 569-591. 27. Lisa Kiang, Tiffany Yip, and Andrew Fuligni, “Multiple Social Identities and Adjustment in Young Adults from Ethnically Diverse Backgrounds,” Journal of Research on Adolescence 18 (2008): 643-670. Tahseen and Cheah: Who Am I? 52 The American Journal of Islamic Social Sciences 35:1 28. Jean S. Phinney and Anthony D. Ong, “Conceptualization and Measurement of Ethnic Identity: Current Status and Future Directions,” Journal of Counsel- ing Psychology 54, no. 3 (2007): 271-81. 29. Jennifer Crocker, Riia Luhtanen, Bruce Blaine, and Stephanie Broadnax, “Collective Self-Esteem and Psychological Well-Being among White, Black, and Asian College Students,” Personality and Social Psychology Bulletin 20, no. 5 (1994): 503-13; Sook K. Kim, “Causal Attribution and Collective Iden- tity among Ethnic and Racial Minority Children and Adolescents,” Disserta- tion Abstracts International (1999) (UMI 9910351). 30. Sirin and Fine, Muslim-American Youth; Maykel Verkuyten, “Ethnic Group Identification and Group Evaluation Among Minority and Majority Groups: Testing the Multiculturalism Hypothesis,” Journal of Personality and Social Psychology 88, no. 1 (2005): 121-38. 31. Carol Ryff. “Psychological Well-Being in Adult Life,” Current Directions in Psychological Science 4, no. 4 (1995): 99-104. 32. Hisham Abu-Rayya, “Ethnic Identity, Ego Identity, and Psychological Well-being among Mixed-ethnic Arab-European Adolescents in Israel,” Brit- ish Journal of Developmental Psychology 24, no. 4 (2006): 669-79; Abdulaziz Aflakseir, “Religiosity, Personal Meaning, and Psychological Well-being: A Study Among Muslim Students in England,” Pakistan Journal of Social and Clinical Psychology 9, no. 2 (2012): 27-31. 33. Fred H Borgen and David C. Barnett, “Applying Cluster Analysis in Counsel- ing Psychology Research,” Journal of Counseling Psychology 34, no. 4 (1987): 456-68. 34. Joe Ward, Jr. (1963), “Hierarchical Grouping to Optimize an Objective Func- tion,” Journal of the American Statistical Association 58 (1963), 236-44. 35. Aldenderfer and Blashfield, Cluster Analysis. 36. Maurice Lorr, Cluster Analysis for Social Scientists (San Francisco, Jossey- Bass, 1983). 37. Glenn W. Milligan, and Martha C. Cooper, “An Examination of Procedures for Determining the Number of Clusters in a Data Set,” Psychometrika 50, no. 2 (1985): 159-79. 38. The dendogram provides a visual representation of the clustering process, as individual cases are combined into clusters. An examination of the den- dogram revealed that a four or six cluster solution might be optimal. Sec- ond, Mojena’s Rule One uses the distribution of within-cluster variance to determine when a jump in coefficients between stages is too large, which suggests that the larger cluster set from the previous stage is a more optimal solution. Based on the most stringent application of Mojena’s rule, the fusion coefficients prior to stage 4 were beyond the acceptable alpha level. This find- ing indicated greater dissimilarity in the clusters, such that fewer than four clusters should not be retained. Finally, the proportional increase in the fu- 53 sion coefficients suggested relatively insignificant jumps in coefficients from all prior stages until stage four, but indicated a significant jump in cluster dissimilarity at stage six. However, the fusion coefficient for stage 6 (i.e. α = 68.91) was further from the lower limit (i.e. α =103.737) than the fusion coefficient for stage 4 (i.e. α =102.33). Thus, based on all three strategies, the six-cluster solution was deemed the most optimal solution and was retained for all subsequent analyses. 39. Ai-Lan Chia, and Catherine L. Costigan, “A Person‐centred Approach to Identifying Acculturation Groups among Chinese Canadians,” International Journal of Psychology 41, no. 5 (2006): 397-412; Madiha Tahseen and Charis- sa S.L. Cheah, “A Multidimensional Examination of the Acculturation and Psychological Functioning of a Sample of Immigrant Chinese Mothers in the US,” International Journal of Behavioral Development 36, no. 6 (2012): 430- 39. 40. Hend Al-Ma’seb, “Acculturation Factors among Arab/Moslem Women Who Live in the Western Culture” (PhD diss., The Ohio State University, 2006) (ProQuest UMI 3226498); Claire Tinker and Andrew Smart, “Constructions of Collective Muslim Identity by Advocates of Muslim Schools in Britain,” Ethnic and Racial Studies 35, no. 4 (2012): 643-63. 41. Branscombe, Schmitt, and Harvey, Group Identification and Well-being. 42. Altaf Husain and F. Ross-Sheriff, “Cultural Competence with Muslim Amer- icans,” in Culturally Competent Practice, ed. Doman Lum, 4th ed. (Belmont, CA: Brooks/Cole Cengage Learning, 2011), 358-390. 43. Pia Rebello Britto and Mona M. Amer, “An Exploration of Cultural Identi- ty Patterns and the Family Context among Arab Muslim Young Adults in America,” Applied Developmental Science 11, no. 3 (2007): 137-50; Krystal Chen and Jane P. Sheldon, “Arab American Emerging Adults’ Bicultural Iden- tity, Acculturation Stress, and Perceptions of Parenting,” Journal of Immigrant & Refugee Studies 10, no. 4 (2012): 438-45; Paul Thomas & Pete Sanderson. “Unwilling Citizens?” Sociology 45, no. 6 (2011): 1028-1044. 44. Huntington, Who Are We? 45. Britto and Amer, Cultural Identity Patterns. 46. John W. Berry, Jean S. Phinney, David L. Sam, and Paul Vedder, “Immigrant Youth: Acculturation, Identity, and Adaptation,” Applied Psychology 55, no. 3 (2006): 303-32. 47. J.E. Marcia, “Identity in Adolescence,” in Handbook of Adolescent Psychology, ed. J. Adelson (New York: Wiley, 1980), 159-87. 48. Saba Ozyurt, “Negotiating Multiple Identities, Constructing Western-Mus- lim Selves in the Netherlands and the United States,” Political Psychology 34, no. 2 (2012): 239-63. 49. Phalet and Güngör, “Moslim in Nederland.” Tahseen and Cheah: Who Am I? 54 The American Journal of Islamic Social Sciences 35:1 50. Abu-Rayya and Abu-Rayya, Identity and Well-being; Sirin and Fine, Mus- lim-American Youth. 51. Khanlou, Koh, and Mill, Cultural Identity. 52. Britto and Amer, Cultural Identity Patterns; Chen and Sheldon, Arab Ameri- can Identity. 53. James E. Cotè and Charles G. Levine, Identity Formation, Agency, and Cul- ture: A Social Psychological Synthesis (Mahwah, NJ: Erlbaum, 2002). 54. Purnima Sundar, “To ‘brown it up’ or to ‘bring down the brown’: Identity and Strategy in Second-Generation, South Asian-Canadian Youth,” Journal of Ethnic and Cultural Diversity in Social Work 17, no. 3 (2008), 251-278. 55. Berry et al., Immigrant Youth Acculturation. 56. Koen Luyckx, Luc Goossens, Bart Soenens, and Wim Beyers, “Unpacking Commitment and Exploration: Validation of an Integrative Model of Ado- lescent Identity Formation,” Journal of Adolescence 29 (2006): 361–378. 57. Ibid.; Marcia, Identity in Adolescence. 58. Berry et al., Immigrant Youth Acculturation. 59. Sirin and Fine, Muslim-American Youth. 60. Tahseen and Cheah, “Multidimensional Examination of Acculturation.” 61. Lena Robinson, “Cultural Identity and Acculturation Preferences Among South Asian Adolescents in Britain: An Exploratory Study,” Children & Soci- ety 23, no. 6 (2009): 442-54. 62. Aldenderfer and Blashfield, Cluster Analysis. 63. American Psychological Association, Guidelines for Providers of Psychologi- cal Services to Ethnic, Linguistic, and Culturally Diverse Populations (Wash- ington, DC: APA, 1993); National Association of Social Workers, Code of Ethics (2008), http://www.socialworkers.org/pubs/code/code.asp. 64. Husain and Ross-Sheriff, Cultural Competence with Muslim Americans; Sun- dar, Identity among South-Asian Youth. 65. Ahmed and Ezzedine, Challenges for Muslim-American Youth. 66. Thea Abu El-Haj, “Global Politics, Dissent and Palestinian-American Iden- tities: Engaging Conflict to Re-invigorate Democratic Education,” in Beyond Silenced Voices: Class, Race and Gender in United States Schools, eds. L. Weis & M. Fine (Albany: SUNY Press, 2005), 119–215.