Prevention of Adolescent Driving Under the Influence

点击翻译按钮获取中文摘要

The purpose of this project is to investigate the effectiveness of a school- and community-based drug prevention program, entitled MVP (Midwestern Prevention Project), in reducing adolescent driving under the influence (DUI) and riding with a driver under the influence (RWDUI). MPP is a social influence-based primary prevention program aimed at modifying the factors related to adolescent gateway drug (tobacco, alcohol, and marijuana) use. The MPP programs were implemented in Kansas City, Kansas, and Kansas City, Missouri, for six years since 1984 and replicated in Indianapolis, Indiana, since 1987. Forty-two middle and junior high schools in Kansas City and 57 schools in Indianapolis participated in the project. The participants were measured either for the longitudinal study or for the cross- sectional study. On average, 1,335 students in Kansas City and 2,519 students in Indianapolis were followed-up each year providing the longitudinal data. For the cross-sectional data collection, an average of 6,373 students in Kansas City and an average of 6,067 students in Indianapolis were measured each year. Prior studies examining the MPP program effects in reducing adolescents' gateway drug use showed successful reduction in smoking, similar reduction in marijuana use, and smaller reduction in alcohol use at both short-term and long-term follow-up assessments. The present project proposes to evaluate the MPP program effects in reducing adolescent DUI and RWDUI, to conduct mediation analysis to determine how the MPP program may have reduced DUI and RWDUI, to conduct moderation analysis to identify subgroups where the program is most or least effective, and to investigate etiology of adolescent DUI and RWDUI to identify the precursors to these behaviors. Advanced statistical methods including latent growth models, mixture models, multilevel models, missing data models, and generalized linear models handling non-normality of data will be applied across multiple waves. In addition, more common statistical methods will be used to extract the maximum information from these data The findings in one project site will be cross-validated in the other site.