SECONDARY ANALYSIS OF SU CAUSAL PROCESSES

描述：（申请人摘要） 苏氏因果关系再分析 本研究的总体目标是：（1）提高我们的认识 成为烟草使用者或滥用者的因果过程， 大麻和其他毒品; 2）改进统计方法 分析纵向数据以探索因果过程。 这项工作将 在现有的七个纵向数据集上进行，每个数据集有四个或 从10到25岁的年轻人中收集更多的数据。 的 具体目标如下： 1. 测试和复制假设的因果过程使用 行为发生和改变的综合理论框架， 三重影响理论（TTI） 具体 关于中介影响、调节效应（包括 性别和种族）和相互因果关系机制（ 预测因子和关于行为经验的反馈效应）。 2. 纵向数据分析的方法有以下五种 领域：物质使用的阶段和水平的概念化， 从一个阶段到另一个阶段的进展，处理随时间变化的预测因素， 数据缺失和损耗的处理，以及建模和测试 因果链中的相互作用（适度调解）。 这项拟议的研究很重要，因为它： 范围（10-25）;能够复制大多数 现有理论中的重要理论;包含更广泛的预测因素 变量比任何其他单一研究;包括来自多个 种族和社会经济多样的样本;利用复杂的 审查调解、主持和反馈的分析技术 关系;具有足够大的组合样本量， 按性别和种族进行分析;提供机会， 多波数据建模的新方法;具有成本效益，因为 六个数据集的存在;并将产生重要影响 未来的预防计划/研究。

DESCRIPTION: (Applicant's Abstract) SECONDARY ANALYSIS OF SU CAUSAL PROCESSES The overall objectives of this research are 1) to improve our understanding of the causal processes involved in becoming a user or abuser of tobacco, marijuana and other drugs; and 2) to improve upon methods of statistical analysis of longitudinal data in exploring causal processes. This work will be conducted on seven existing longitudinal data sets, each with four or more waves of data collected from youth between the ages of 10 and 25. The specific objectives and aims are as follows: 1. Test and replicate hypotheses regarding causal processes using the framework of an integrative theory of behavior onset and change developed by these investigators, the Theory of Triadic Influence (TTI). Specific hypotheses concern mediating influences, moderating effects (including of gender and ethnicity) and reciprocal causation mechanisms (both among predictors and regarding feedback effects of behavioral experience). 2. Extend methods of longitudinal data analysis in the following five areas: conceptualizing stages and levels of substance use, modeling of progression from stage to stage, dealing with time-varying predictors and endogeneity handling missing data and attrition, and modeling and testing interactions in causal chains (moderated mediation). The proposed study is important because it: encompasses the high risk age range (10-25); enables replication of tests and contrasts of the most important of the existing theories; incorporates a wider range of predictor variables than any other single study; includes data from multiple ethnically and socioeconomically diverse samples; utilizes sophisticated analytical techniques to examine mediating, moderating and feedback relationships; has large enough combined sample sizes to enable complex analyses by gender and ethnicity; provides opportunities to develop and test new methodology for modeling of multi-wave data; is cost-efficient because of the existence of the six data sets; and will have important implications for future prevention programs/research.