Measuring Executive Function During Early Childhood: The Utility of Direct Assessments, Teacher Ratings, and Group-Based Tasks

Abstract

Among the many factors contributing to children’s development, executive function (EF) skills have received a considerable amount of attention in recent years, given their role in developmental outcomes, such as academic achievement and behavior problems. Yet, much of our understanding of EF development is grounded in experimental studies that rely on highly controlled, laboratory-based measures, usually administered one-on-one to individual children. While these techniques have provided insight into the development of EF, they may not account for important contextual factors that might influence the ways in which children employ EF in important learning environments such as the classroom. EF and related cognitive and self-regulatory skills often play out in group settings, with natural distractors and social interactions with peers and teachers. Increasing the ecological sensitivity of early EF measures has the potential to enhance our understanding of how EF manifests in a naturalistic classroom setting and how it relates to specific classroom behavior and academic outcomes. Therefore, the goals of present dissertation were to 1) develop and validate a new set of EF measures, specifically designed to capture young children’s EF in a dynamic social context with peers and distractors 2) compare these measures to direct assessments, and to teacher ratings of children's EF, and 3) examine the degree to which performance on EF tasks across measurement types is related to children's academic achievement during kindergarten. Results from study 1) provide psychometric evidence that a group-based paradigm is a valid approach to studying EF processes in socially demanding contexts. Findings from study 2) revealed only modest correlations between individually assessed EF components and components derived from teacher ratings. Relations among teacher-rated EF and the group-based tasks were larger in magnitude and mapped onto corresponding sub-components across both measurement types, which might suggest that the social context exerts a similar demand on the way children employ EF skills in a classroom and group setting. Results from study 3) revealed that patterns of predictions to math and reading skills across measurement contexts were similar in significance, direction and magnitude of effect. Finally, although teacher-reported EF factors were statistically significant predictors of both math and reading achievement in kindergarten, interpretations were limited due to the high levels of multicollinearity between teacher-reported EF factors. Contributions to developmental theory are discussed.