Do Young Children’s Executive Functions Vary by Context and Measurement Approach?

• In Event: 1-144 - Executive Functions in Flux: Contextualising our Approaches for the Real World

Integrative Statement

Children’s executive functions (EFs) are often measured with lab-based tasks in one-to-one, distraction-free environments. Lab-based tasks ignore real-life challenges of a classroom environment, where children face a variety of distractors. Teacher reports of EFs achieve greater ecological validity, but are subjective in nature (Mason, Gunersel, & Ney, 2014). At best, lab-based EF measures correlate moderately with teacher reports of EFs (Toplak, West, & Stanovich, 2013). Another approach to measuring EFs is to use classroom observations. This method relies on trained researchers’ ratings of children’s behavior in an ecologically valid setting. The present study compares these three EF assessment methods, to shed light on the context-dependent nature of EFs and the ecological validity of each method.

We present analyses from a subsample of 96 children in the UK (54.2% girls; Mage = 4.70 years) from 15 classrooms attending schools that cater primarily for children from low income backgrounds (full sample n = 200+ children; 32 classrooms; 14 schools). To examine how EFs might appear to vary across contexts, children’s EFs were measured with three measurement approaches: lab-based assessments, teacher reports, and classroom observations. Lab-based assessments were administered in a one-to-one format in a quiet space in children’s schools. We included three externally-driven tasks, meaning children follow set rules to produce set responses, and two self-directed tasks, meaning children could produce a range of responses that were correct (White, Burgess, & Hill, 2009). Teachers reported on children’s school readiness using the Brief Early Skills and Support Index (BESSI; Hughes et al., 2015), including the Behavioural Adjustment sub-scale that captures EF-like behaviors. Finally, children’s EFs were also observed for 4-7 minutes in a child-led and teacher-led situation in their classroom. Their behavior was coded on the 16-items of the Regulation Related Skills Measure (RRSM; McCoy et al., 2017).

Children’s scores on externally-driven and self-directed lab-based assessments were significantly correlated (r(91) = .29, p = .005, controlling for age, non-verbal intelligence and language). Teacher reports of EFs were positively, though weakly, associated with children’s observed EFs in child-led (r(79) = .24 p = .03) and teacher-led (r(88) = .34, p = .001) classroom situations. Notably, children’s observed EFs in child-led and teacher-led situations were not related. There were no other significant correlations.

Findings suggest EFs vary across contexts and within individuals. Whereas teachers’ perspective on EFs, as evidenced by their reports, seemed to tap in to observed classroom behaviors, these same behaviors were unrelated to a range of lab-based assessments of EFs.

The interplay between measures and theory is discussed. When deriving models of EF development, it no longer seems appropriate to rely solely on lab-based assessments.

Authors

• Janina Eberhart, University of Cambridge
  Presenting Author

• Hayley Gains, University of Cambridge
  Non-Presenting Author

• Sara Baker, University of Cambridge
  Non-Presenting Author