Drinking From Ages 13-21 in College/Non-College Samples

Texas Tech University, Lubbock TX

Investigators

Abstract

DESCRIPTION (provided by applicant): This one-year R21 application seeks to advance the literature on college-student drinking via a secondary analysis of the Family and Adolescent Study (FAAS), a six-year longitudinal investigation conducted in metropolitan Buffalo, New York. The sample consists of adolescents who were 13-16 years of age at the first wave of annual assessments (six-wave N = 506). The focus of the proposed secondary analysis will be on changes (expected to be increases) in heavy drinking, as students advance through junior high and high school into college. The FAAS was conducted with a representative, general-population sample, thus containing a large number of individuals who did not go to college, who will serve as an age-matched comparison group to respondents who did. Three specific aims of the proposed study are to: (1) replicate recent longitudinal results from national datasets showing increases in heavy drinking as individuals progress from high school to college, and a relative lack of change over the same time period among individuals who did not go to college; (2) extend these findings by examining individuals starting at a much younger age (13) than in these national studies; and (3) extend the range of variables used to predict increased drinking during this age span using a theoretical framework focusing on factors that may predispose students toward partying/peer-socialization contexts in school and on actual participation in these activities. Respondents' patterns of change in heavy drinking will be examined primarily via latent growth models. Multiple-group modeling will be done, to test predictive models separately within college- and non-college-bound respondents. Given the staggered nature of the original research design - with one set of respondents' six-year participation taking place from ages 13-18, another's from 14-19, etc. - a cohort sequential design will be used to simulate a nine-wave design in which a single cohort progresses from ages 13 to 21. Findings of significant predictive relationships between respondents' psychosocial characteristics during adolescence and increased drinking up through the college/young adult years - if obtained - will potentially aid the development of preventive interventions by: (1) strengthening the justification for beginning such interventions at early ages; (2) providing an indication of which variables, specifically, have such long-term predictive efficacy; and (3) determining which variables' abilities to predict later heavy drinking are common to college- and non-college-bound individuals, and which are unique to just one of these groups.