We evaluated the potential impact of participant attrition on the results of the primary analyses. The sample included in the analyses (N = 1,160) was compared to the drinkers who dropped out of the study prior to the final assessment (N = 868) on key variables assessed during senior year of high school (age, gender, ethnicity, family history, impulsivity, heavy drinking, and problems). AO and AI were collected at the final assessment, by which point maximum attrition had occurred; therefore, analyses examining differences in these variables across samples were not possible.
