Comparing the factor structures of cognitive measures of executive function and parent ratings of executive function in a mixed clinical group
MetadataShow full item record
The construct of executive function (EF) has been extensively researched in recent years due to its proposed role in a variety of clinical disorders. However, there remains contention between researchers regarding the definition and components of the construct (Eslinger, 1996; Packwood, Hodgetts, & Tremblay, 2011.) The current study sought to explore the construct of EF by investigating the factor structures of performance-based (i.e., cognitive) measures of EF as well as parent ratings of EF (i.e., behavioral measures). Data were culled from an archival database of neuropsychological case studies submitted as part of the KIDS, Inc School Neuropsychology Post-Graduate Certification Program. The database was narrowed using specific exclusion criteria. The final dataset consisted of 176 participants between the ages of 8 and 16. Two exploratory factor analyses were completed using full information maximum likelihood to account for missing data. The first analysis examined behavioral measures of EF and revealed a five factor structure. Factors were named Externalizing, Internalizing/Self-Regulation, Adaptive, Metacognition, and ADHD. The second analysis examined cognitive measures of EF and revealed a two factor structure. Factors were named Shifting and Reasoning. Pearson product moment correlations were then computed to investigate the relationships between the factors obtained during each of the factor analyses. Correlations between behavioral and cognitive factors were weak. Lastly, five regression analyses (i.e., one for each of the five behavioral factors) were completed for each cognitive variable in order to determine whether the behavioral variables making up each factor predicted performance on cognitive variables. The Adaptive and Internalizing factors appeared to best predict performance on the D-KEFS variables, while the Adaptive, Internalizing, and Metacognition factors best predicted performance on the NEPSY-2 variables. Overall, the factors were poor predictors of performance on the WJ III COG NU. The results of this study provide additional evidence that the construct of EF is highly complex and thus difficult to define and effectively measure. Furthermore, behavioral and cognitive measures of EF should not be considered interchangeable; instead, they should be considered separate but key components of neuropsychological assessment.