The second meditational hypothesis was that alcohol’s effects on the ERN are at least partially derived from reduction of negative affect. In support of this hypothesis, Figure 8 shows that the indirect effect of alcohol on the ERN via NA change was significant (standardized estimate = .21, p < .01). The model also showed two other significant indirect effects. First, the effect of alcohol on posterror adjustment via the ERN was significant (standardized estimate = .21, p < .01), indicating that the reduction of evaluative control reflected in the ERN accounts for significant variability in alcohol’s impairment of posterror adjustment. Second, the model shows that significant additional variability in posterror adjustment is accounted for by inclusion of the NA change variable in the model, as indicated by the compound path from alcohol to NA change, from NA change to the ERN, and from the ERN to posterror adjustment (standardized estimate = 0.10, p = .02). No other process variables in the model provided significant explanatory power. However, it should be noted that the indirect effect linking the control beverage variable
