Importance of negative consequences in the statistical prediction of self-reported problematic drinking among college students

Purpose. To investigate the replicable association between negative consequences and self-acknowledged problem drinking among college students. Method. This study uses multiple random samples from five years of annual alcohol and other drug web-based anonymous surveys of undergraduate students at a large public university. Each sample includes at least 1,500 students 18-22 years old, with a total of over 9,000 students; data includes demographic variables, quantity and frequency measures of drinking, 17 negative consequences and self-reported problem drinking. Statistical methods include factor analysis, scale reliabililty, and logistic regression on self-reported problem drinking using samples from two years. Replicability of the regression equation was tested on the remaining three years of data. Results. Negative consequences factored into two scales: one of seven relatively frequent consequences and one of ten less frequent (but overtly more serious) consequences. Each scale had an internal reliability of at least .7 in the combined sample. Controlling for sex, race and average number of drinks per week, both scales of negative consequences significantly increase the probability of problem drinking, with the scale of relatively frequent consequences having the strongest relationship to problem drinking. Conclusions. Dangerous drinking, that is, drinking patterns associated with a high probability of negative consequences, is a national problem among college students. The impact of dangerous drinking on students is significant, but several studies have demonstrated that students are often resistant to acknowledging the seriousness of negative consequences. Objective information relating negative consequences to self-acknowledged problem drinking could assist intervention efforts.