I think I can't: Drink refusal self-efficacy as a mediator of the relationship between self-reported drinking identity and alcohol use

• We evaluated drinking identity as a moderator of drinking.
• Drinking identity and drink-refusal self-efficacy interacted to predict drinking.
• Four measurement models for latent variables were tested.

This study examined the relationship between self-reported drinking identity (SRDI), defined as how closely individuals believe drinking is a crucial aspect of their identity (Conner, Warren, Close, & Sparks, 1999), and alcohol use by considering drink-refusal self-efficacy (DRSE) as a potential mediator. Based on previous findings, we expected that SRDI would be negatively associated with DRSE and positively associated with drinking, and that DRSE would be negatively linked with drinking. Further, we expected that DRSE would mediate the association between SRDI and drinking. Participants included 1069 undergraduate students (M age = 22.93 years, SD = 6.29, 76.25% female) from a large southern university who completed computer-based study materials. Gender was associated with SRDI, each of the DRSE subscales, and drinking, indicating that males report greater SRDI, lower DRSE, and increased alcohol consumption. Consistent with expectations, SRDI was negatively linked with DRSE and positively linked with drinking. DRSE subscales were negatively associated with drinking. Further, four measurement models for latent variables were tested for SRDI and each of the three DRSE subscales. Results showed that the emotional relief and social subscales of DRSE mediated the association between SRDI and drinking, however this mediating relationship did not emerge for the opportunistic subscale. Implications of these results are discussed.