The primary aim of this research was to test potential mechanisms of the effects of alcohol on neurophysiological manifestations of error processing and impairment in self-regulatory control that previously have been observed (e.g., Ridderinkhof et al., 2002) or theorized (e.g., Holroyd & Yeung, 2003; Yeung & Cohen, 2006; Yeung, Ralph, & Nieuwenhuis, 2007). Addressing this aim requires examination of the relations among the variables to test for potential mediation of outcomes (e.g., ERN amplitude, posterror adjustment) by theorized mechanisms (e.g., error recognition, affect changes). To do so, we constructed a structural equation model and tested the significance of indirect effects related to hypothesized mechanisms. As discussed by Kenny (2009), using a structural equation model to test for potential mediation is advantageous because it permits tests of multiple mediators simultaneously, which can show whether a given indirect or mediated effect is independent of other indirect or mediated effects on the outcome of interest (see also MacKinnon, Fairchild, & Fritz, 2007).
