ADVANCED STATISTICAL TECHNIQUES FOR PREVENTION RESEARCH

点击翻译按钮获取中文摘要

In order to understand the complexity of how drug use prevention interventions may work among adolescents, more sophisticated statistical techniques will be needed. Four relevant projects have been proposed to offer broader methodological perspectives in prevention research, especially in the evaluation of intervention effects: (1) Application of the multi-sample approach with regression models and structural equation models to more adequately evaluate intervention effects on adolescents' substance use behavior; (2) Application of models incorporating developmental growth to evaluate prevention effects with adolescents' progression in substance use; (3) Application of multilevel models to investigate the impacts of school context on adolescents' substance use; and (4) Investigation of the impacts of participant attrition in prevention research. Monte Carlo (computer simulation) study and empirical study will be conducted to investigate the performance of these models applied to prevention research, particularly in the area of evaluation of prevention programs. The applicant's supporting institution is the Institute of Health Promotion and Disease Prevention Research (IPR), within the Department of Preventive Medicine, School of Medicine, University of Southern California. The principal goal of IPR is the prevention of disease. To accomplish this goal, the institute has an extensive portfolio of research protocols designed to promote primary prevention among children and adolescents. IPR research focuses on multi-component, comprehensive community approaches to prevention. IPR developed the Midwestern Prevention Project (MPP), which is a school-based substance use prevention project targeting adolescents in two Midwestern cities. The longitudinal data obtained from MPP will be made available to the applicant to perform evaluation and empirical testing of the proposed advanced statistical methods.