Collegiate alcohol consumption and academic performance.

OBJECTIVE Although studies consistently have found a negative bivariate association between alcohol use and academic performance among college students, some research suggests that this association largely results from student differences at matriculation. The present study examines this relationship while controlling for key background factors. METHOD Personal interview surveys were conducted for four consecutive semesters with random samples of students at a small, liberal arts college, resulting in a combined sample of 754 (392 women). The interviews measured alcohol consumption, gender, race, athletic status, academic class, parents' education and income, and frequency of attending off-campus parties; and 94% of the sample granted permission to obtain grade point average (GPA), high school class rank, and Scholastic Aptitude Test (SAT) scores from official college records. RESULTS The amount of alcohol consumed correlated significantly with GPA (r=-.26). Ordinary least squares regressions showed that gender and partying accounted for 43% of variation in alcohol consumption, and that academic class and parents' income had positive indirect effects on consumption. SAT score and class rank were the strongest predictors of GPA, but alcohol consumption remained significant when these and other variables were controlled (beta = -.24 when controlling for SAT, and beta = -.14 when controlling for both SAT and class rank in a smaller, biased subsample). CONCLUSIONS The disparity in findings between this and previous research was explained in terms of differences in type of institution studied, which suggests the need to consider the college context and the interaction of college and individual factors in studies of college drinking.