Validity of the K-Bit and a WISC-III short form with a sample of adolescent young offenders

Abstract

The present study investigated the concurrent validity and testretest reliability of the Kaufman Brief Intelligence Test (K-BIT; Kaufman & Kaufman, 1990) by comparing it to the full scale administration of the Wechsler Intelligence Scale for Children third edition (WISC-III; Wechsler, 1991) in a clinical trial with 35 young offenders. Using a test-retest methodology, the concurrent validity of a four subtest short form of the WISC-III (SF4) was also assessed. Results of this study show that while the K-BIT is a reliable measure, it did not perform well in terms of validity criteria typically applied to short form estimates of intelligence. While mean Full Scale IQ (FSIQ) and Performance IQ (PIQ) were not significantly different, the KBIT overestimated Verbal IQ (VIQ) by an average of five points. Correlation coefficients between the K-BIT and WISC-III were significant but not impressive and only 40% to 50% of the IQ estimates were within WISC-III 95% confidence limits. Qn the other hand, SF4 performed reasonably well as an estimate of FSIQ, with a high correlation coefficient and a low misclassification rate. When administered first SF4 overestimated FSIQ by three IQ points, however when administered following the full scale, SF4 overestimated FSIQ by six IQ points.