Hypothesis 2 predicted that the SRE-5 score that relates to early drinking periods would correlate more robustly with maximum drinks and the number of alcohol problems in the younger generation whose current age was closer to the time of first drinking than was true for probands. This prediction was supported by higher offspring zero-order correlations for SRE-5 to maximum drinks and the number of alcohol problems in Table 2 than was observed for probands in Table 1. Results supporting Hypothesis 2 were also present in Table 5 where higher relationships of SRE-5 to alcohol outcomes were generally observed in offspring compared with probands, even after considering covariates. However, that pattern was not observed in the regression analyses in Table 6 where the contribution of SRE-5 to the regression analyses was more robust in the probands. The absence of significant contributions of SRE-5 in the regression analyses for the alcohol outcomes for offspring in Table 6 is likely to reflect the higher correlations among their SRE measures for offspring shown in Table 2 compared with correlations among SRE values for probands
