Psychology of Addictive Behaviors, Vol 37(3), May 2023, 390-401; doi:10.1037/adb0000900
Objective: Level of response (LOR) to alcohol is associated with several alcohol-related risk factors and outcomes. However, existing self-report measures of LOR have important limitations. For example, the Self-Rating of the Effects of Alcohol Scale assesses a limited range of alcohol-related effects. Although the Alcohol Sensitivity Questionnaire (ASQ) samples a broader range of effects, it uses different probes across effects, confounding type of effect with method variation associated with the use of different probes. Focusing on the ASQ, we systematically evaluate variation in estimated LOR as a function of how number of drinks to achieve an effect is probed. Our approach addresses a major limitation of existing LOR measures which fail to account for sensitivity variability across drinking occasions. Method: This study randomized 732 adult drinkers into one of four versions of the ASQ that assessed sensitivity to 15 alcohol-related effects, systematically varying the follow-up probes. Results: Accounting for (a) the minimum number of drinks consumed before feeling an effect and (b) the maximum number of drinks consumed without feeling an effect for all effects is superior to the original ASQ approach in predicting relevant outcomes. Conclusions: Assessments of sensitivity should probe for minimum and maximum number of drinks across each of the effects. If impractical to probe for both, consistently probing for maximum number of drinks is desirable. (PsycInfo Database Record (c) 2023 APA, all rights reserved)