Moderators affecting outcomes in a preventive, family -based intervention for at -risk young adolescents

Celine F Provini, Fordham University


This study built upon a previously conducted demonstration trial that supported the general efficacy of Multidimensional Family Prevention (MDFP) for strengthening protective factors related to young adolescents' avoidance of alcohol, tobacco, and other drug (ATOD) use and delinquent behavior. The present investigation re-analyzed demonstration-trial data, aiming to answer a more specific question regarding efficacy—that is, whether participant characteristics (adolescent symptomatology, parental symptomatology, and parental education), termed “amenability-to-intervention” variables, moderated positive, MDFP-related changes in targeted outcomes including adolescent self-competence, adolescent school orientation, and family cohesion. The current study also examined pretest scores on these outcomes as additional potential moderators of the same outcomes. Thus, the study sought to assist future MDFP implementers in targeting appropriate populations for intervention, as well as tailoring the intervention for particular types of participants. The sample consisted of 124 young adolescents (mean age of 12.5 years) and their functional parents. Participants resided in inner-city Philadelphia, a nationally designated high-risk area, and adolescents were recruited for the study based on screening results indicating moderate to high risk for problem behaviors. Virtually all families were African American. Participants were assessed on study variables at two time points (pretest and posttest), and hierarchical regression was used to analyze the data. Results indicated that adolescent symptomatology had a small moderating effect—approaching statistical significance—on family cohesion; adolescents with higher pretest symptomatology levels experienced smaller MDFP-related increases in family cohesion than did adolescents with lower pretest symptomatology levels. Adolescents with higher pretest symptomatology levels also experienced greater declines in self-competence over the course of the study than did adolescents with lower pretest symptomatology levels; however, this effect was not intervention-related, as it applied to both MDFP and control groups. Pretest scores on outcome variables did not emerge as moderators of the same outcomes. Therefore, only one moderating relation emerged—adolescent symptomatology acted as a nearly significant amenability-to-intervention variable regarding MDFP's effect on family cohesion. This finding does not warrant routine assessment of adolescent symptomatology prior to implementation of preventive family interventions, although it does suggest that prevention programmers may wish to investigate other potential moderators relevant to particular interventions.

Subject Area

Developmental psychology

Recommended Citation

Provini, Celine F, "Moderators affecting outcomes in a preventive, family -based intervention for at -risk young adolescents" (2003). ETD Collection for Fordham University. AAI3077257.