VULNERABILITY TO ADOLESCENT DRUG USE AND DEVIANCE

问题行为理论认为人格系统是不可或缺的

Problem behavior theory posits an integral role for the personality system in the etiology of drug use (DU). Tests of the relative efficacy of personality as a predictor of DU have primarily emphasized "direct- effects" approaches rather than dynamic mediational models. Two essential problems are associated with this view: (l) during adolescence personality may be one of many factors that mediate the effects of important social and cognitive competencies on later DU; and (2) most evaluations of personality features linked to DU have emphasized their "risk-engendering" properties, emphasizing less their potential role as protective factors. Two lines of research suggest a new role for elements of the personality structure as protective factors: (l) empirical confirmation that personality factors may mitigate or offset risk conditions; and examination of protective factors in concert with risk factors demonstrate their utility in predicting uniquely to DU. Given the intransigence of many risk factors, a different primary prevention tactic suggests enhancing inoculatory effects in preventing initiation to DU. Additionally, problem behavior theory has focused mainly on the early stages of DU; however, empirical findings point toward different sets of psychosocial precursors for early versus later stages of DU. In effect, prevention programs may need to broaden their focus to include other developmental patterns including maintenance, exacerbation, and cessation of DU. The current study proposes to examine the relative efficacy of several key components of the Jessors' personality structure as predictors of DU (in addition to testing consequence models). The study design consists of a secondary analysis of data delimited to include control students only participating in a five-year longitudinal school-based intervention program. Measures included in the study span a wide range of perspectives of adolescent DU and include: elements of the Jessors' personality system (motivational-instigation, perceived control, and perceived belief structure), psychological well-being, cognitive expectancies (attitudes), normative expectations, perceived peer and adult DU, and psychosocial measures related to social and cognitive competence. Distinction of developmental pathways for various patterns of DU will utilize hierarchical regression strategies and multivariate variance decomposition appropriate for repeated measures. Logistic regression and discriminant function analysis will be used to distinguish empirically categorical group membership. Structural equation modeling will be used to contrast alternative theoretical models and conduct multiple group comparisons based on demographic characteristics including age, gender, and ethnicity. Results will forge a stronger conceptual and empirical bridge between etiology theories and prevention.