Psychometric properties of the UngDOK

Abstract

Aim: This article describes and discusses the Swedish UngDOK interview and its psychometric properties. Method: The study is based on empirical data from 1633 intake interviews collected by 15 units in ten cities and focused primarily on the two central sections of intake form: alcohol and drug use and mental health. The statistical analyses concern internal consistency, test–retest reliability, discriminant validity and internal non-response. Results: The reliability of AUDIT-C and the mental health domain was good with regard to both internal consistency and test–retest. The test–retest values were generally satisfactory, except for frequency of drug use and association with peers who use drugs. The discriminant validity shows that the interview clearly distinguishes adolescents with more profound problems from a group with milder problems and that a minor degree of inconsistency and non-response bias may occur in empirical material based upon self-reported information. Conclusion: The study showed that the psychometric properties of the UngDOK interview are generally satisfactory and may be regarded as a valuable option for practices engaged in treating adolescents with substance-use problems.