Examining the psychometric properties of the WISC-IV in low and normal IQ samples

Abstract

Background The Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) is one of the leading instruments in general ability assessments; however, findings regarding its factor structure are still controversial. Methods In this study, the single-factor, correlated four-factor, higher-order factor, and bifactor models for the 10 basic tests of the WISC-IV were examined on systematically controlled samples with low (IQ <= 79; N = 141) and normal scores (IQ > 80; N = 222). Results Our research findings revealed that only the bifactor and correlated four-factor models showed good fit in both sample groups; however, measurement invariance was provided only by the correlated four-factor model. Conclusions In general, our findings indicate that the use of the full-scale intelligence quotient (FSIQ) score should be used with caution in both low and normal IQ groups and that it would be more appropriate to prefer subscale index scores, especially in educational and clinical assessments.