Sociology Ph.D. Dissertations


School Racial and Ethnic Composition Effect on Academic Achievement of Latino Adolescents

Date of Award


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Sociology/Population Studies

First Advisor

Jennifer Van Hook


Although the U.S. student population has grown increasingly diverse both in terms of ethnicity and immigrant generational status since the late 1980s, schools have become more racially and ethnically segregated. Data also reveal that Latinos, the nation’s largest minority, have become increasingly segregated over the last 30 years, with their segregation levels surpassing those of blacks. In this dissertation, I investigate the effects of school racial composition on Latino adolescents’ academic achievement. The primary reason for focusing on Latinos adolescents is that they consistently account for the highest high school dropout rate among the nation’s major ethnic groups. Previous research suggests that academic achievement is a function of both individual and family level characteristics. Using the National Longitudinal Study of Adolescent Health (Add Health) data I examine the interplay of school racial and socioeconomic composition, school social capital, family social capital, ethnic origin, and immigrant generational status on measures of school success, such as school grades and standardized test scores, while controlling for individual (e.g., sex, age) and family (e.g., family structure, SES) factors. The longitudinal Add Health data possess a hierarchical structure such that the individual-level factors are viewed as nested within the school-level factors. Hierarchical linear modeling is used as an appropriate statistical procedure for examining these nested data. I found that school racial composition has little, if any, effect on Latino students’ academic achievement, but school socioeconomic composition does. Importantly, family social capital is likely to mitigate harmful influences of attending a low-SES school. I also found school social capital, as measured by peer network homogeneity and density, to be positively associated with Latino achievement. More than any other Latinos, Cuban-American adolescents were shown to have higher achievement in schools with more homogeneous and denser peer networks. Apart from that, I did not find any significant differences between ethnic origin and academic achievement. Similarly, beyond the first generation, immigrant generation was not associated with Latino achievement. However, the first-generation immigrant youth was found to have significantly lower AHPVT scores, but higher GPAs than native Latino adolescents.