| The present study investigated the mediating effect of implicit alcohol associations and moderating effect of temptation coping strategies on the relationship between affective decision-making and alcohol use among adolescents. The design and interpretation of results followed the theoretically distinct, yet complementary dual process models of adolescent cognition and addictive behavior (Steinberg, 2007; Thush & Weirs, 2007). A total of 100 adolescents from a northwest urban high school were recruited to participate in a single session of assessment. After removal of 22 participants through data screening, the final sample ranged in age from 14--18 years (M=15.6; 54% female); ethnic representation included 60.4% Caucasian, 17.2% Multiracial, 9.6% Asian/Pacific Islander, 7.3% Hispanic, and 5.5% Black/African-American. Results indicated partial support for a developmentally informed, dual process model of alcohol use. Supporting H1, results of regression analyses indicated that affective decision-making predicted alcohol use (beta = -.33, t = -3.09, p < .01). Affective decision-making did not predict implicit alcohol associations, and thus, H2 was not supported (beta = .00, t = .17, p = .86). Supporting H3, implicit alcohol associations were found to predict alcohol use (beta = .36, t = 3.35, p < .01). Results of a Preacher and Hayes (2008) mediation analysis with bootstrapped resampling methods did not indicate support for the mediating effect of implicit alcohol associations on the relationship between affective decision-making and alcohol use (H4); the indirect effect was a1b1+a 2b2 = -.0281, CI = -.015 to .007. Utilizing a critical value of p < .10 to determine a statistically significant moderation effect (Aguinis, 2004), results of regression analyses indicated that temptation coping significantly moderated the relationship between affective decision-making and alcohol use (beta = .001, t = 1.96, p = .05), supporting H5. Post-hoc multiple regression was conducted to investigate unique independent effects of study variables. Following the proposed pathways, alcohol use was significantly predicted by affective decision-making, implicit alcohol associations, and temptation coping F(3, 73) = 30.06, p < .01. Together, these predictors accounted for 70% of variance in alcohol use. Findings replicate and extend previous research on adolescent drinking and may improve screening of at-risk youth. |
