The analyses addressing the 4 hypotheses were analyzed through 4 stages. Those included the following: (1) zero-order correlations among alcohol challenges (for probands), measures of Time 1 SRE, and Time 1 and Time 2 past 5-year values for maximum drinks per occasion and number of alcohol problems; (2) evaluations of potential characteristics that might need to be controlled when evaluating how the LR measures relate to the 2 key alcohol outcomes including age, sex, race, education, marital history, presence of children, impulsivity, and sensation seeking; (3) results regarding how different LR measures related to cross-sectional and prospective alcohol outcomes while controlling for covariates; and (4) hierarchical multiple regression analyses where covariates and all relevant LR measures except for the composite SRE-T were entered into the same regression analyses for each of the key outcomes.
