Construct validity of the Wechsler Intelligence Scale for Children - Fourth UK Edition with a referred Irish sample: Wechsler and Cattell-Horn-Carroll model comparisons with 15 subtests

Autor: Rebecca Good, Marley W. Watkins, Gary L. Canivez, Kate James, Trevor James
Rok vydání: 2017
Předmět:
Zdroj: British Journal of Educational Psychology. 87:383-407
ISSN: 0007-0998
DOI: 10.1111/bjep.12155
Popis: Background Irish educational psychologists frequently use the Wechsler Intelligence Scale for Children – Fourth UK Edition (WISC–IVUK; Wechsler, 2004, Wechsler Intelligence Scale for Children-Fourth UK Edition, London, UK, Harcourt Assessment) in clinical assessments of children with learning difficulties. Unfortunately, reliability and validity studies of the WISC–IVUK standardization sample have not yet been reported. Watkins et al. (2013, International Journal of School and Educational Psychology, 1, 102) found support for a bifactor structure with a large sample (N = 794) of Irish children who were administered the 10 WISC–IVUK core subtests in clinical assessments of learning difficulties and dominance of general intelligence. Because only 10 subtests were available, Cattell–Horn–Carroll (CHC; McGrew, 1997, 2005, Contemporary intellectual assessment: Theories, tests, and issues, New York, NY: Guilford; Schneider & McGrew, 2012, Contemporary intellectual assessment: Theories, tests, and issues, New York, NY, Guilford Press) models could not be tested and compared. Aim, Sample and Method The present study utilized confirmatory factor analyses to test the latent factor structure of the WISC–IVUK with a sample of 245 Irish children administered all 15 WISC–IVUK subtests in evaluations assessing learning difficulties in order to examine CHC- and Wechsler-based models. One through five, oblique first-order factor models and higher order versus bifactor models were examined and compared using CFA. Results Meaningful differences in fit statistics were not observed between the Wechsler and CHC representations of higher-order or bifactor models. In all four structures, general intelligence accounted for the largest portions of explained common variance, whereas group factors accounted for small to miniscule portions of explained common variance. Omega-hierarchical subscale coefficients indicated that unit-weighted composites that would be generated by WISC–IVUK group factors (Wechsler or CHC) would contain little unique variance and thus be of little value. Conclusion These results were similar to those from other investigations, further demonstrating the replication of the WISC–IV factor structure across cultures and the importance of focusing primary interpretation on the FSIQ.
Databáze: OpenAIRE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje