This dissertation consists of two studies that examined the development of same-ethnic friendship preferences over the course of middle school and its predictors and consequences. Both studies draw from a longitudinal school-based study of about 6,000 ethnically diverse early adolescents’ social and psychological adjustment in 26 schools that varied in ethnic diversity. I employed a measure of same-ethnic friendship preference used in past research that accounts for the probability of having a same-ethnic friend, given the opportunities for forming both same- and cross-ethnic friendships in school. In Study 1, I examined the developmental trajectory of same-ethnic friendship preference and considered the compositional and organizational features of schools and classrooms that lead to these preferences. Results from latent growth curve models revealed that same-ethnic friendship preference increased over the course of middle school and that these friendship preferences were shaped by both the school and classroom ethnic context, above and beyond availability. At the start of middle school when friendship networks are beginning to form, having fewer same-ethnic peers in schools that were less ethnically diverse predicted steeper increases in same-ethnic friendship preference. This interactive effect of school ethnic diversity and same-ethnic peers in school did not predict changes in friendship preferences. Rather, greater school ethnic diversity predicted steeper increases in same-ethnic friendship preference over time. Taking a more dynamic approach to understanding the ethnic composition, I also examined how youth are organized within schools in their academic classes. Results revealed more complex and longitudinal associations. African American and Asian youth, groups that are strongly academically stereotyped, who were underrepresented in honors classes showed steeper increases in same-ethnic friendship over time. In Study 2, I evaluated the interplay of same-ethnic friendship preference and ethnic identity and intergroup attitudes over the course of middle school, and whether these associations depended on the school ethnic context and ethnicity. The results revealed that for ethnic minority youth, same-ethnic friendship preference at the start of middle school predicted less steep decreases in ethnic identity over time. This effect did not vary by proportion of same-ethnic peers in school. The effect of ethnic identity on changes in same-ethnic friendship preferences was weaker, and only found among Asian youth when they had few same-ethnic peers in school. The longitudinal associations between same-ethnic friendship preference and attitudes favoring one’s ingroup were weaker. Although friendships and attitudes were positively correlated at the start of middle school, their slopes indicated that they did not develop together over time. Additionally, a friendship selection effect was found such that attitudes favoring one’s ingroup predicted steeper increases in same-ethnic friendship preference over time. However, friendships did not predict greater biases in attitudes over time.

Taken together, by considering the school structural and organizational features that lead to same-ethnic friendship preference, as well as the longitudinal associations between friendship preference, ethnic identity, and intergroup attitudes over the course of middle school, this dissertation identified how both formal and informal aspects of schools can lead to greater ethnic boundaries between youth, and the factors that can diminish those boundaries. Implications for prejudice reduction interventions and creating more inclusive school environments are discussed.

The present cross-sectional study used structural equation modeling (SEM) to examine the links between early adolescents' perceptions of the school racial climate, sense of belonging, academic identity, and academic performance in the spring of 6th grade. It further explored whether the strength of model relationships differed based on adolescents' ethnicity. The sample included 3,969 (52% female) students from 19 ethnically diverse urban middle schools. Based on self-reported ethnicity, 12.9% were African-American/Black, 14.7% East/Southeast Asian, 17.9% White, and 37.8% Latino. Findings revealed that within multiethnic schools, experiencing supportive racial norms has important consequences for ethnic minority adolescents' academic outcomes. This was partially mediated by youths' sense of belonging to school and academic identity. A stronger relationship between racial climate and school belonging was found for Asian, Latino, and White youth compared to African-American peers. The relationship between racial climate and academic identity was strongest for Latino and African-American youth. Finally, the relationship between racial climate and academic outcomes was nonsignificant for White adolescents but significant for all ethnic minority youth. Implications for future research are discussed.