Descriptive statistics and zero-order correlations for the study variables are shown for the sample in Table 1. Preliminary analyses indicated that the high school conduct problems and baseline and follow-up alcohol use variables were positively skewed (skew = 1.63, 5.34, and 5.40, respectively). As described above, logarithm transformations were applied in order to normalize variables so that their skew was less than |1| prior to running any analyses. All variables were positively intercorrelated (rs = 0.06 – 0.69), with the exception that the tetrachoric correlation between parental alcohol problems and dating several people was non-significant.
