EVALUATION OF ADOLESCENT OFFENDERS WITH SUBSTANCE-ABUSE - VALIDATION OF THE SASSI WITH CONDUCT-DISORDERED YOUTH

The adolescent version of the Substance Abuse Subtle Screening Invento ry (SASSI-A; Miller, 1990) was developed to assess unacknowledged subs tance abuse. Despite its widespread use in clinical and forensic setti ngs, the SASSI-A has not been cross-validated. The current investigati on examined its effectiveness in classifying 317 adolescent offenders on a unit of dually diagnosed inpatients. Although the sample was limi ted in its number of nonusers, the SASSI-A had an unacceptably high nu mber of false positives (68.4%) but was moderately effective at classi fying nonadmitting alcohol and drug users (75.6%). As evidence of crit erion-related validity, elevations on the SASSI-A scales had low to mo derate correlations with interview-based data on impairment related to substance abuse. However, its scales appeared to be significantly aff ected by ethnicity, even when level of impairment was a covariate. It appears, then, that the SASSI-A (a) should not be employed to classify adolescents as chemically dependent and (b) has a circumscribed role in screening for suspected substance abuse.