Between-task exploratory analysis
These observations raise an important possibility: The fact that alcohol had no effect on mean SPN amplitude in the Regularity task could be partly due the fact that the participants had marginally (although non-significantly) higher AUDIT scores in the Regularity task (Mean AUDIT in Oddball task = 9.65 (SD =4.84); Mean AUDIT in Regularity task = 12.00 (SD = 3.96), t (50) = 1.911, p = .062, ds = 0.530).
Regression analysis supports this possibility (Figure 10). The DV was alcohol-induced change in SPN amplitude. When entered alone, AUDIT explained 17.8% of variance in the DV (R2 = 0.178, F (1,50) = 10.818, p = .002). Inclusion of Task to the model explained little additional variance (R2 change = 0.024, F (1,49) = 1.466, p = .232). Given this analysis, we cannot confidently claim that Task has an independent effect on alcohol induced changes in SPN amplitude, beyond the effect of AUDIT.
Unfortunately, we did not obtain AUDIT scores from the 13 participants in the pilot study. It could be that the pilot participants were relatively heavy drinkers, explaining why alcohol enhanced the SPN in this sample. This is plausible, considering that the maximum possible AUDIT score is 40, and the pivot from SPN reduction to enhancement happened at around 15 in the Oddball task.