B - 74 Using Cattell-Horn-Carroll Theory to Investigate the Structure of Executive Function in the Delis-Kaplan Executive Function System (D-KEFS).

Abstract

To evaluate a series of published executive function factor models for the Delis-Kaplan Executive Function System (D-KEFS) and contrast these models with an a priori three-factor model based on Cattell-Horn-Carroll (CHC) theory. This was a cross-sectional study using the 16 main achievement scores from the D-KEFS standardization sample, which includes three separate age-cohorts. We aimed to replicate previously reported factor models for the D-KEFS in the 20-to-49-year cohort (N = 361) using the approaches described in the original published studies. We also sought to use confirmatory factor analysis to contrast these models with an a priori three-factor model based on CHC theory. The previously reported models failed to replicate, despite a large number of iterations, suggesting that these models are not plausible for the D-KEFS. Initial observations indicated an admissible but poorly fitting three-factor CHC model. However model fit was improved by adding correlated residuals for scores obtained from similar tests, reflecting method effects. These methods effects, derived from one sample, replicated in the other two standardization samples. The final CHC model demonstrated good to excellent fit for the D-KEFS in all three samples, and displayed strong metric invariance across the three age-cohorts. The results highlight the importance of replication in neuropsychological assessment, and show that D-KEFS scores can be interpreted in line with CHC constructs of processing speed, fluid reasoning, and retrieval fluency.