The factor structure of executive functions measured with electrophysiological correlates: An event-related potential analysis

Abstract

The three-factor model of executive functions is widely employed in cognitive control research. However, recent studies have revealed psychometric problems with commonly used difference scores in behavioral measures of executive functions. Examining behavioral scores, several studies were unable to find a coherent factor structure for executive functions or identify significant individual differences in specific executive function abilities. These findings have raised questions about the utility of established measurement scores for executive functions. Our study sought to reassess the three-factor model proposed by Miyake et al. (2000), employing event-related potentials (ERPs) from electroencephalography (EEG) as a means to directly probe underlying cognitive processes, leveraging the EEG’s high temporal resolution. We conducted an analysis of the factor structure of the three executive functions (updating, shifting, and inhibition) in a sample of 148 participants. We employed Bayesian structural equation models to examine the relationships between the mean amplitudes of the N2 and P3 components, obtained from a battery of nine executive function tasks. Our results indicate that amplitudes of the ERP components measured in executive function tasks almost exclusively represent variance related to general processes rather than executive function-specific variance. Notably, no task demonstrated variance uniquely attributable to individual differences in executive function processes added through experimental manipulations. These results cast doubt on the validity of current executive function tasks in accurately reflecting individual differences in these processes.