Respondent misreporting of drug use in self-reports: Social desirability and other correlates

To assess possible validity problems associated with self-reported drug use, drug use forecasting (DUF) data collected in 1995 from the Birmingham (AL) catchment area were used to examine the statistical correlates of arrestees misreporting their drug use by comparing self-reported drug use involving (1) marijuana, (2) cocaine, and (3) both cocaine and marijuana with results of urinalysis tests performed on the arrestees. We used the social desirability theory to explain why some arrestees misreported using drugs while others did not. Multivariate analyses revealed, as predicted by the theory, differences between the two groups based on drugs consumed and personal, legal, and other factors. Our results have implications for the validity of larger-scale self-reported estimates of the extent and nature of drug use in the U.S. that must be addressed before realistic policy choices can occur.