This doctoral dissertation, which includes two distinct studies, focuses on the development of educational expectations and educational utility values in African American adolescents. Guiding frameworks for the project include expectancy-value theory (Wigfield & Eccles, 2000), theory on adolescent development (Steinberg, 2001), Cooley’s (1902) symbolic interactionist theory, and the concepts of risk and resilience (Luthar et al., 2000). Data for the study are drawn from the African American subsample (N = 876) of the Maryland Adolescent Development in Multiple Contexts Study (MADICS) (Eccles, 1997), and span the six year period between the time during which youth were enrolled in Grade 7 and one year after high school graduation.

In the first study, multiple regression and logistic regression models demonstrated that reflected appraisals of parents and teachers were related to African American adolescents’ educational utility values and expectations for future educational attainment during Grade 7 and again during Grade 11. The strength of the relation between reflected appraisals of parents and youths’ educational utility values decreased between Grade 7 and Grade 11, whereas the strength of the relation between reflected appraisals of teachers and utility values increased over time. The magnitude of associations between reflected appraisals and expectations/values did not change during this period. In addition, findings suggested that affiliation with achievement-oriented peers may buffer youth from the harmful impact of low reflected appraisals of teachers during Grade 7, and from the harmful impact of low reflected appraisals of parents during Grades 7 and 11. Finally, reflected appraisals of parents and teachers during Grade 11 were associated with high school graduation status at levels that were marginally significant. The relation between reflected appraisals of parents and completion of high school appeared to be mediated by youths’ educational expectations.

In the second study, latent growth models were used to examine trajectories of educational expectations and utility values between Grade 7 and Grade 11, with a special focus on gender differences. Boys’ and girls’ expectations and values were statistically equivalent during Grade 7. On average, expectations did not change across the time period examined, although the trajectory for boys was significantly less positive than the trajectory for girls. In addition, the sample average for educational utility values decreased across time in a manner that was the same for girls as for boys. Although the hypothesized gender difference in college participation did not materialize, logistic regression models revealed that educational expectations and educational utility values uniquely contributed to the variability in adolescents’ college participation status one year after high school graduation.