To test for the effects of alcohol use disorder (AUD) and cannabis use disorder (CUD), we utilized 2 × 2 factorial ANCOVA models, with AUD and CUD as predictor variables, each with two levels (use disorder versus no use disorder). Additionally, to account for demographic differences, we added age and sex as covariates into the models. This design was utilized on behavioral performance (accuracy and reaction time), as well as neural activity in each time–frequency window of interest to interrogate the independent main effects of AUD and CUD. Our design was specifically aimed at testing the independent main effects of AUD and CUD to maximize statistical power. That is, the main effect of AUD compares all participants with AUD against all participants without AUD. Because the effects of AUD and CUD are examined in the same model, the effect of the other use disorder is controlled for, effectively controlling for comorbid use disorders. Interaction terms were not included due to the relatively small sample size of the subgroups.
