The Validity of Different Measures of Automatic Alcohol Action Tendencies

• Published in: Psychology of addictive behaviors 2015-03, Vol.29 (1), p.225-230
• Main Authors: Kersbergen, Inge, Woud, Marcella L., Field, Matt
• Format: Article
• Language: English
• Subjects: Action; Addictive behaviors; Adolescent; Adult; Alcohol Abuse; Alcohol Drinking - psychology; Alcohol Use; Alcoholism - psychology; Approach Avoidance; Attention; Automatic; Automatism - psychology; Avoidance Learning; Behavioral responses; Choice Behavior; Classification; Drinking behavior; Female; Human; Humans; Latency; Male; Methodological problems; Pattern Recognition, Visual; Prediction; Psychology; Reaction Time; Relatedness; Risk Factors; Stimulus; Symbolism; Task Analysis; Tobacco; Young Adult
• ISSN: 0893-164X; 1939-1501
• DOI: 10.1037/adb0000009

Previous studies have demonstrated that automatic alcohol action tendencies are related to alcohol consumption and hazardous drinking. These action tendencies are measured with reaction time tasks in which the latency to make an approach response to alcohol pictures is compared with the latency to make an avoidance response. In the literature, 4 different tasks have been used, and these tasks differ on whether alcohol is a relevant (R) or irrelevant (IR) feature for categorization and on whether participants must make a symbolic approach response (stimulus-response compatibility [SRC] tasks) or an overt behavioral response (approach avoidance tasks [AAT]) to the pictures. Previous studies have shown positive correlations between measures of action tendencies and hazardous drinking and weekly alcohol consumption. However, results have been inconsistent and the different measures have not been directly compared with each other. Therefore, it is unclear which task is the best predictor of hazardous drinking and alcohol consumption. In the present study, 80 participants completed all 4 measures of action tendencies (i.e., R-SRC, IR-SRC, R-AAT, and IR-AAT) and measures of alcohol consumption and hazardous drinking. Stepwise regressions showed that the R-SRC and R-AAT were the only significant predictors of hazardous drinking, whereas the R-AAT was the only reliable predictor of alcohol consumption. Our results confirm that drinking behavior is positively correlated with automatic alcohol approach tendencies, but only if alcohol-relatedness is the relevant feature for categorization. Theoretical implications and methodological issues are discussed.