Using path analysis to understand executive function organization in preschool children.

Abstract

There continues to be no consensus definition of executive functions. One way to understand different executive function components is to study abilities at their emergence, that is, early in development, and use advanced statistical methods to understand the interrelations among executive processes. However, to fully determine the constructs of interest, these methods often require complete data on a large battery of tasks, which are difficult to obtain with young children. Path analysis is an alternative statistical technique that requires only a single measure of each construct, yet still allows researchers to investigate complex relations among measures, to compare nested models, and to compare model fit across groups. Therefore, 117 preschool children (ages 2 years 8 months to 6 years 0 months) completed several executive function tasks. Path analysis was used to determine the relations between complex problem solving and working memory, inhibition, and set shifting processes. The best-fitting model included paths from working memory and inhibition to problem solving, and a correlation between working memory and inhibition. Interestingly, in younger children, inhibition was the strongest predictor of problem solving, whereas working memory contributed more strongly in older children. Suggestions for useful statistical methods to investigate the relations among executive functions in children are discussed.