PCA-based estimates of N1 sink amplitude were highly comparable across paradigms, as revealed by correlations computed for participants who provided data from both paradigms (AH, n = 14, .81 ≤ r ≤ .98; NH, n = 30, .87 ≤ r ≤ .92; HC, n = 42, .89 ≤ r ≤ .92; all p < .001). To increase statistical power, data stemming from the two paradigms were therefore combined into a single N1 sink estimate by using either the available N1 sink estimate or the pooled mean across paradigms, exploiting that PCA factor scores represent a standardized measure (the overall mean across participants, tasks, and sites is zero with a standard deviation of 1). As expected, the repeated measures ANOVA for the combined sample (AH, n = 26; NH, n = 49; HC, n = 46) confirmed the findings for each paradigm, however, at a more robust significance level (group main effect, F[2,118] = 4.26, p = .02, η2p = .067; between group contrasts: AH vs. NH, F[1,118] = 6.55, p = .01, AH vs. HC, F[1,118] = 7.44, p = .007, NH vs. HC, F[1,118] < 1.0, n.s.).
