UC Irvine

This dissertation presents three studies that examine the role of person characteristics, primarily noncognitive skills and gender, in the person-context relations of out-of-school experiences and school-related functioning. Using structural equation modeling, I investigate noncognitive skills in three ways: as selection factors predicting participation in out-of-school contexts, as moderators of the associations between out-of-school activity participation and outcomes, and as mediators of the associations between out-of-school activity participation and outcomes. Multiple-group structural equation models were used to test for gender moderation simultaneously in the analysis of pathway models. The first study modeled the cross-predictive relations between person characteristics (noncognitive skills, behavioral problems, and school grades) and intensity of participation in out-of-school contexts at age 15 and end of high school. Key findings include a positive bidirectional relationship between school grades and intensity of participation in organized activities, and a gender moderated effect in which for boys, but not girls, higher level of noncognitive skills and behavioral problems predicted increased unsupervised time, and more unsupervised time predicted decreased school grades at the end of high school. The second study modeled two person characteristics (defiance and college expectations) as moderators of relationships between intensity of participation in out-of-school experiences over two years of middle school and school-related and risk-taking outcomes. Moderation effects were consistent with compensatory and dual risk hypotheses, such that organized activities were most compensatory for high defiant boys’ school attendance, and unsupervised activities were most risky for high defiant boys’ school attendance and high defiant boys’ and girls’ drug use. The last study of this dissertation modeled a noncognitive skills latent variable (indicated by work orientation, self-reliance, and self-identity) as a mediator of longitudinal associations between consistency of organized activity participation from Kindergarten to Grade 5 and school grades in high school. Noncognitive skills significantly mediated these longitudinal associations for girls. Only a direct, non-mediated model was found for boys. While research often conceptualizes person characteristics as demographic and control variables, the results of this dissertation emphasize the importance of accounting for the selection, moderating, and mediating roles of person characteristics, including gender and non-cognitive competencies, in the developmental process.