Validation and comparison of three different versions of the Drinking Motives Questionnaire: a network analysis approach.

Abstract

Alcohol is widely used, but motives for consumption vary. Since drinking motives are proximal predictors of initiation and escalation of alcohol use, accurate measurement of these motives is crucial. The aim of this study was to assess the psychometric properties - including factor structure, internal consistency, and test-retest reliability - of the 12-item, 18-item, and original 20-item versions of the Drinking Motives Questionnaire (DMQ). Additionally, this study aimed to examine the associations between drinking motives and alcohol use outcomes using network analysis, providing insights into the direct and conditional relationships between different drinking motives and their real-life outcomes. A total of 434 adults (77.4% female) completed the baseline questionnaires and 130 (85.3% female) participated in a follow-up approximately two weeks later. Results showed that the subscales of the DMQ had good internal consistency ( .80) regardless of version. Test-retest reliability was acceptable to good (rs= .71 to rs= 83) but was higher for social and enhancement than for conformity and coping motives. Confirmatory factor analysis showed that the 12-item short form outperformed the original 20-item version and the more recently proposed 18-item version. Partial correlation networks including DMQ subscales and alcohol outcomes revealed consistent positive association between social motives and binge drinking, alcohol use quantity per drink occasion and enhancement motives, while frequency of use was primarily associated with coping motives. However, while associations amongst drinking motives were identical in the 12- and 18-item version, these differed from the 20-item version. This study demonstrates how network models can be valuable tools in the validation and comparison of questionnaires and shows that the 12-item DMQ could be a good alternative for the often used longer forms.