Derek Heim

Introduction Children's early experiences with alcohol inform the development of alcohol‐related beliefs which are known to predict alcohol consumption during the critical stage of adolescence. Yet, there has been considerably less research into these alcohol‐related cognitions in low‐ and middle‐income countries (LMIC) and existing measures of these beliefs are highly reflective of Western contexts, which may not be fully appropriate for use in LMICs. The aim is to ascertain the construct validity of the Alcohol Expectancies Questionnaire (AEQ) in a non‐Western sample. Methods A cross‐sectional diagnostic accuracy study involving 500 adolescents aged 10 to 18 years randomly selected from the database of the Navrongo Health and Demographic Surveillance System. Participants were administered the locally back translated version of the 34‐item AEQ. Confirmatory factor analysis using the lavaan package in R was conducted to generate indices for the factor structure of the AEQ. Results Confirmatory factor analyses showed that while groupings of positive and negative expectancies were similar to those observed when expectancies have been assessed previously in Western studies, these formed a single ‘alcohol expectancy’ factor. Questions relating to positive tension reduction and negative physical expectancies showed inconsistent responses in this study. Discussion and Conclusions Commonly used tools for the assessment of alcohol expectancies may not be suitable for use in Ghana, possibly owing to their development and validation in Western contexts. These findings have implications for the assessment of alcohol‐related beliefs in LMIC settings and begin to map out a research agenda to develop more contextually and culturally attune alcohol assessments.