Investigating the structure of the WJ-III Cognitive at school age

During its development, the Woodcock-Johnson, Third Edition Cognitive (WJ-III Cognitive; McGrew & Woodcock, 2001) was never subjected to structural analysis using exploratory and higher order factor analyses. Instead, confirmatory factor analyses were conducted on separate sets of WJ-III correlation matrices, yielding a seven-factor model across all age ranges. To see whether the structure holds for the WJ-III Cognitive, currently recognized best practice exploratory factor analysis (EFA) procedures were applied to two school-aged correlation matrices (ages 9-13; 14-19) from the normative sample. Using EFA and higher order factor analysis, four factors emerged at age 9 to 13 and three factors emerged at age 14 to 19. The results of this analysis indicated a robust manifestation of general intelligence (g) that exceeded the variance attributed to the lower order factors. An additional analysis was conducted that disregarded factor extraction rules and forced the seven-factor fit. The resulting solution was only partially aligned (i.e., Gc, Ga, and Gsm) with the theoretical structure posited in the WJ-III Technical Manual. Surprisingly, this study represents the first time to my knowledge that the WJ-III Cognitive has been subjected to EFA analyses given the instrument's significant use by practitioners and that it has served as the initial evidentiary base for Cattell-Horn-Carroll theory.