The full model (constructed using Mplus version 5.21; Muthén & Muthén, 2007) is presented in Figure 8. The beverage group variable was dummy coded using the placebo group as the comparison for the alcohol and control groups. Thus, effects presented in the model for Alcohol and Control represent effects of those beverages relative to the placebo beverage. This approach high-lights both pharmacological effects (alcohol vs. placebo) as well as any expectancy-related effects (control vs. placebo). Two primary meditational hypotheses were investigated. First, the error-detection view of the ERN (Coles, Scheffers, & Holroyd, 2001) suggests that alcohol’s reduction of ERN amplitude, and subsequent impairment of posterror adjustment, reflect impaired awareness of errors (Ridderinkhof et al., 2002). Although the response accuracy judgment findings presented previously have suggested that error awareness was intact in the alcohol group, it remains possible that variability in this outcome could play a role in alcohol’s effects on posterror adjustment. In addition, the Pe component, associated in previous work with error awareness (Hester et al., 2005; Nieuwenhuis et al., 2001), was also reduced by alcohol, suggesting some support
