Measuring cognitive ability fairly: Measurement invariance of the KABC-II NU across parent education levels

This study examined whether measurement invariance is supported for the Kaufman Assessment Battery for Children, Second Edition Normative Update (KABC-II NU), across parental education groups, a common proxy for socioeconomic status. Establishing measurement invariance allows for comparing scores across demographic groups and provides evidence for the assumption that the KABC-II NU measures intelligence similarly regardless of the examinee's parents' education. Data from the KABC-II NU U.S. standardization sample were analyzed in this study, which included 507 youth ages 7-18. The sample was divided into groups by the highest level of education attained by their parent(s): Grade 11 or less (n = 55), high school graduate or equivalent (n = 109), some college/tech certification/associate's degree (n = 173), and bachelor's degree (n = 165). Data were analyzed using multigroup confirmatory factor analysis. Configural invariance and full metric invariance were supported. Partial intercept invariance was achieved by freeing the intercept for Block Counting. Residual invariance and equality of the first-order disturbances across groups were supported. The variance of g did not meaningfully differ across parental education, suggesting a similar degree of variability in general intelligence (width of the normal curve) across groups. The estimated latent broad ability means showed that students with less educated parents demonstrate a relative weakness in verbal comprehension. These findings provide additional validity evidence for the KABC-II NU and suggest the test is not biased against youth from diverse socioeconomic backgrounds. (PsycInfo Database Record (c) 2026 APA, all rights reserved).