The problem of substance use among adolescents is of great concern to researchers, policy makers, and the public in terms of health, education, and social consequences. Because most data regarding substance use and its correlates have been collected in natural settings, conclusions about the causal pathways to substance use and abuse have been constrained. The goal of the training portion of this grant is to enhance. both the Pi's skills in the statistical techniques for furthering knowledge of the development and prevention of youth drug use and the Pi's knowledge base in issues relating to youth substance use with a focus on ethnic and gender differences. The goal of the research program is to apply contemporary and new methods to improve understanding about the normative developmental trajectories of substance use during adolescence and the causal mechanisms in those trajectories across a range of gender and ethnic groups. The first general aim of the research is to identify general and subpopulation-specific models of the prediction and development of normative adolescent substance use, using innovative trajectory and mixture modeling methods to analyze data from the NIDA-funded National Survey of Parents and Youth-. Given the normative trajectories in this large dataset for different ethnic and gender groups, we will then identify the ethnic- and gender-specificity of established predictors of adolescent substance use. The second general aim is to test hypotheses of mediated causal effects of an intervention on adolescent substance use in high-risk youth, using data from the NIMH-funded Fast Track project, a long-term intervention program and field experiment intended to reduce youth conduct problems. We will first test the replicability of the findings from the first general aim, then - using techniques developed in the training program - evaluate potential causal pathways and mediating relations linking the Fast Track intervention to changes in adolescent substance use. The NSPY and Fast Track datasets represent unique opportunities to study the development of substance use in both experimental and non-experimental contexts. We will use recent and novel advances in statistical modeling, in which the PI will be trained, to advance our understanding of the complex role of ethnicity, gender, and other constructs in substance use prevention.
Showing the most recent 10 out of 26 publications
