Given the criteria for inclusion in the current analyses, all participants completed at least the first and final assessments. However, of the 2,245 participants who completed assessment 1, nearly 32% dropped out of the study by the final assessment. To evaluate the potential impact of attrition on the results of the primary analyses, we used chi-square analyses and independent sample t-tests to compare baseline values on study variables between study completers (N = 1,160) and the sample of drinkers who dropped out of the study before the final assessment (N = 868). Chi-square tests indicated that those included in the analyses were more likely to be female participants (66.0% vs. 53.8%), χ2(1 df) = 31.17, p < 0.001, and Caucasians (58.7% vs. 51.8%), χ2(7 df) = 61.03, p < 0.001, relative to those who were not included. Included participants were also less impulsive than those who dropped out, t(df = 1,944) = −2.67, p = 0.008. No differences were observed for age, frequency of heavy drinking, or problems during senior year of high school. Suggesting that our participants' drinking behavior
