Hayes, 2008). In addition, the macro allows for the inclusion of covariates as well as potential mediators in the analysis model. We included the demographic variables of age, gender (0 = male, 1 = female), and race (0 = white/caucasian, 1 = black or other) in all models tested, as well as a measure of family history of alcoholism (family history density from the FTQ). The latter is both a strong predictor of alcohol dependence and a risk factor for childhood maltreatment, as children of alcohol dependent parents are more likely to experience parent-child conflict (Hill et al., 1994). Psychiatric comorbidity (i.e., lifetime diagnosis of a mood, anxiety, or other substance use disorder, 0 = no, 1 = yes) was also included in the covariates for AD subjects (there was little to no psychiatric comorbidity in the CON subjects). Indirect effects were estimated via bootstrap analysis using 5000 randomly generated samples. Mediation is established if the 95% bias-corrected confidence interval for the indirect parameter estimate does not contain zero. All statistical analyses were conducted using SAS version 9.2 (SAS Institute, Cary, NC). The level of statistical significance was set at p < 0.05 for all tests.
