Validation of the Adolescent Drinking Expectancy Questionnaire and development of a short form

Abstract

Aims

This study aimed to validate the Drinking Expectancy Questionnaire—Adolescent version (DEQ-A) in a large adolescent sample and to develop and validate a brief measure, the Drinking Expectancy Questionnaire—Shortened Adolescent version (DEQ-SA). Design and Methods

Cross-sectional survey of secondary school students (n = 2357, aged 13–16, M = 14.66 years, SD = 0.60). Students completed the DEQ-A in school, and measures of alcohol consumption including the Alcohol Use Disorders Identification Test—Consumption. The data were randomly split, and Exploratory Factor Analysis was performed using subsample 1, and Confirmatory Factor Analysis and reliability and validity testing were performed using subsample 2. Results

The 24-item DEQ-A was successfully reduced to 12 items (DEQ-SA) without compromising psychometric properties. The DEQ-A and the DEQ-SA both demonstrated adequate-to-good fit to the data and very good internal reliability. The DEQ-A and DEQ-SA explained 20 and 18% of the variance in alcohol consumption. Adolescents who drank endorsed more positive alcohol expectancies, whereas alcohol-naïve adolescents scored higher on negative alcohol expectancies. As the DEQ-SA comprises two subscales of the DEQ-A, the endorsement rates are applicable to both scales. Discussion and Conclusions

The DEQ-A and the short form of this scale developed in this study (DEQ-SA) show good reliability, internal structure and account for a large proportion of variance in alcohol consumption. Both scales can assist in targeting cognitive change processes within tailored alcohol prevention and treatment approaches, and investigating hypothesised mechanisms of change. The DEQ-SA is recommended for more time-limited environments.