As shown in Table 1, the ability of the SRE to predict alcohol outcomes overlaps with the predictive value of the alcohol challenge. Running regressions where only SRE LR or SHAS was entered predicting T2 drinking, then a regression where SRE LR and SHAS were simultaneously entered predicting T2 drinking, then running a regression where SRE LR was entered first followed by SHAS, and finally a regression where SHAS was entered first followed by SRE LR generated a series of R2’s. These allowed for the calculation of the unique variance that SRE LR shared with T2 drinks (R2 = .126), the unique variance that SHAS shared with T2 drinks (R2 = .027), and the T2 drinks variance mutually shared by SRE LR and SHAS (R2 = .041). These relationships are demonstrated graphically in Figure 1 in a Venn diagram. Overall, as shown by the squares of the correlation in Table 1, the SRE-based LR explained 16.7% of the variance in predicting T2 drinks, with, as shown in Figure 1, 12.6% explained by the SRE independent of the alcohol challenge LR.
