Comparative psychometric study of a range of hazardous drinking measures administered online in a youth population.

Abstract

To compare the psychometric performance of a range of existing alcohol measures when data are collected online with young people, and thereby to gain insights into the reliability and validity of this mode of data collection. One hundred and sixty-seven U.K. resident young people aged 16-24 who had drunk alcohol within the past week participated in a cross-sectional psychometric study with a test-retest reliability component. Eight hazardous drinking measures were used: the alcohol use disorders identification test (AUDIT) summary instrument and dedicated assessments of consumption (timeline follow-back and diary-format recall of alcohol drunk in the last 7 days), dependence (Leeds dependence questionnaire and severity of dependence scale) and problems (Rutgers alcohol problem index, alcohol problems scale and academic role expectations and alcohol scale). Internal consistency and test-retest correlation statistics were generally satisfactory, providing evidence of reliability. Validation data obtained in principal components analyses, investigation of the correlation matrix and in a multiple regression model of total AUDIT score were also supportive of the online use of these measures. Evidence was weakest for the alcohol problems scale. A range of hazardous drinking measures exhibit sound psychometric properties when administered online. Further comparative study of the relationships between different measures is needed.