Accuracy of detained juveniles self report of substance use on the Massachusetts Youth Screening Instrument-Version 2

Self-report measures are the mainstay of psychological assessment. The reliability and validity of self-report measures are always a concern. In this study, detained juveniles' self-report accuracy is examined using a comparison of biological measures and self-report on the MAYSI-2. The sample consisted of 741 juveniles detained in the Midwest between the ages of 12 and 18 over a six month period. Out of this group 663 individuals had complete files to include both biological measures and self-report results. The study found that 37.1% of the sample tested positive on biological measures for illegal substance use (Figure 3) and 28.8% of the sample indicated on item 23 of the MAYSI-2 that they had gotten in trouble while either high or drunk (Figure 1). Approximately 16% of the sample fell in the caution area of the substance use scale as opposed to the 33% in the normative sample, and 2.7% of the sample fell in the warning area as opposed to 10% in the normative sample (Figure 2). Comparing the mean substance use scale of the sample to the mean of the nonnative sample produced a Cohen's d of .39, indicating the considerable difference between the two samples. The sensitivity of item 23 was 41.5% and the specificity of item 23 was 78.7% (Table 12). Along with a low positive and moderate negative predictive value, the data indicated that item 23 was not a good indicator of substance use. Due to the denial of substance use, 22% of the sample was identified as false negatives. Age did not appear to have a significant influence on accuracy.