Promoting positive youth development : using within-regression analysis to estimate group process growth curves as mediators of intervention change in multi-problem youth

This study extended the knowledge of and examined the moderation and mediation effects of contextual interpersonal exogenous and endogenous variables on intervention response. The first analyses consisted of Repeated Measures Multivariate Analysis of Variance (RMANOVAs) and the second analyses consisted of estimating linear growth curves using within-subject regression analysis. The results of the outcome and moderation analyses identified ethnicity as a plausible moderator of intervention response. Specifically, African Americans demonstrated significant higher psychosocial development in intimacy compared to Hispanics. The mediation analyses, which estimated linear growth curves using within- subject regression analysis, identified a potentially influential mediator, that is, group impact, as a causal factor contributing to the positive psychosocial development of identity. In summary, the current study furthered examined the psychosocial development of identity and intimacy. Secondly, it demonstrated the influential moderation and mediation effects on intervention response. And thirdly, it expanded the methodological procedures in a university-community based intervention