Structural Validity of the Woodcock Johnson III Cognitive in a Referred Sample by Strickland, Tracy, Ph.D., Arizona State University, 2012 , 85 pages; AAT 3505630
The structural validity of the WJ-III Cognitive was investigated using the GIA-Extended Battery test scores of 529, six-to-thirteen-year-old students referred for a psychoeducational evaluation. The results of an exploratory factor analysis revealed 11 of the 14 tests loaded on their expected factors. For the factors Gc , Gf , Gs , and Gv , both tests associated with the factor loaded highly; for Gsm , Glr , and Ga , only one test associated with each factor loaded highly. Obtained congruence coefficients supported the similarity between the factors Gs , Gf , Gc , Glr , and Gv for the current referred sample and the normative factor structure. Gsm and Ga were not found to be similar. The WJ-III Cognitive structure established in the normative sample was not fully replicated in this referred sample. The Schmid-Leiman orthogonalization procedure identified a higher-order factor structure with a second-order, general ability factor, g , which accounted for approximately 38.4% of common variance and 23.1% of total variance among the seven, first-order factors. However, g accounted for more variance in both associated tests for only the orthogonal first-order factor Gf . In contrast, the Gc and Gs factors accounted for more variance than the general factor for both of their respective tests. The Gsm , Glr , Ga , and Gv factors accounted for more variance than g for one of the two tests associated with each factor. The outcome indicates Gc , Gf , Gs , and Gv were supported and thus are likely factors that can be utilized in assessment while Gsm , Glr , and Gr were not supported by this study. Additionally, results indicate that interpretation of the WJ-III scores should not ignore the global ability factor.
Posted using BlogPress from Kevin McGrew's iPad