Evaluation of data screening methods in surveys of adolescents' drug use.

This study evaluated methods to identify adolescents who inaccurately report drug use in school surveys. Two samples of seventh graders (N=2,710) completed a battery of questionnaires including measures of drug use. Subjects classified as random responders, subjects with a relatively large number of inconsistent responses, and subjects who reported using a fictitious drug all reported higher rates of drug use than did other subjects. For random responders and fictitious drug users, the rate of reported drug use was inconsistent with other subject data. Implications of these findings for the use of exclusionary criteria in surveys of adolescent drug use are discussed