Table 1 shows four separate ordinary least squares (OLS) regressions predicting, alternately, N2 or P3 peak with theta and delta to investigate the relationship between time-domain and TF representations of N2 and P3. For predicting the grand average N2 scores, grand average theta and delta scores accounted for 46.72% of the variance, and critically, both accounted for unique variance. For the P3 grand average measure, theta and delta explained 83.84% of the variance, and again, each TF component accounted for unique variance in P3. For the no-go minus go difference models, theta and delta difference scores accounted for 37.60% of variance in the N2 component scores, with both theta and delta having unique and near equal contributions to the model. Finally, theta and delta difference scores accounted for 74.10% of the variance in the no-go minus go P3 difference scores, with both accounting for unique variance in the model. Regression assumptions of residual normality, independence of residuals, and homoscedasticity were satisfied.1
