Higher-Order Factor Structures for the WISC-IV: Implications for Neuropsychological Test Interpretation

Abstract

Factor-analytic studies support a hierarchical four-factor model for the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) with a prominent general, third-order factor. However, there is substantial disagreement on which type of higher-order model best fits the data and how different models should guide test interpretation in clinical practice, with many studies concluding interpretation should primarily be focused on general indicators of intelligence. We performed a series of confirmatory factor analyses with the WISC-IV standardization sample (N = 2,200, ages 6–16 years) to examine model fit and reexamined models used to support test interpretation at the general level. Consistent with previous research, bifactor models were difficult to identify; however, compared with bifactor and hierarchical models, the correlated factors model with no general higher-order factor provided the best fit to the data. Results from this study support the basic four-factor model specified in the WISC-IV technical manual, with test interpretation primarily focused at the factor level, rather than the general level suggested in previous studies.

Key words:

• assessment
• factor analysis
• intelligence
• WISC-IV

Notes

WISC-IV = Wechsler Intelligence Scale for Children-Fourth Edition; PR = Perceptual Reasoning; WM = Working Memory; PS = Processing Speed; VC = Verbal Comprehension; TLI = Tucker Lewis Index; CFI = Bentler's Comparative Fit Index; PNFI = parsimonious normed fit indicator; PCFI = parsimony comparative fit index; RMSEA = root mean square error of approximation; AIC = Akaike information criterion.