Risky drinking of alcohol in Sweden: A randomized population survey comparing web- and paper-based self-reports

AIMS - This study compared data quality and response rates for the Alcohol Use Disorders Identification Test (AUDIT) in a web-based versus in a paper-and-pencil format survey. SETTING - Subjects were randomised to one of two parallel samples, one web-based and the other paper-based. Data were collected during 2009. RESULTS - The web-based format yielded a lower response rate compared to the paper version (26.2% vs. 53.6%), internal consistency was quite similar (0.82 vs. 0.77), while the mean AUDIT scores were higher in the web-based format for both men and women. Conclusions - Future studies should focus on methods for combining different administration methods in order to maximize response rates.