Behavior Ratings and Direct Assessment of Executive Function as Predictors for Later Academic Achievement

• In Event: 1-197 - Assessment and Development of Executive Function in Early Childhood

Integrative Statement

Introduction
A large number of studies has identified executive function (EF) as a significant predictor of academic achievement across childhood (Best, Miller & Jones, 2009). Some scholars have administered direct assessment tasks of EF and others have applied EF rating measures. Previous studies report moderate correlations between direct assessment and adult ratings of EF (McAuley, Chen, Goos, Schachar & Crosbie, 2010). Taken together, these findings raise the question of the respective relevance of the two types of measures for the prediction of academic achievement, which is also the aim of the present study.
Method
N=170 kindergarten children (45.6% girls; Mage=6.10 years, SD=.13) were assessed around their sixth birthday on tasks measuring EF subcomponents. Updating was assessed using the complex span task of the German Working Memory Test Battery for Children aged 5 to 12 Years (AGTB 5-12; Hasselhorn et al., 2012). The border version of the Dimensional Change Card Sort (DCCS; Zelazo, 2006) was administered as a measure of shifting and the Head-Toes-Knees-Shoulder task (HTKS; Ponitz, McClelland, Jewkes, Connor, Farris & Morrison, 2008) provided an index of inhibition. In addition, the children’s kindergarten teachers rated the children’s EF using a newly developed questionnaire shortly before school entry (at the age of six). At the end of Year 1 of primary school children’s reading, spelling and mathematical performance was assessed using standardized school achievement tests based on the German primary school curriculum.
Results & Discussion
Separate hierarchical regression analyses were calculated for reading, spelling and mathematical performance. For each model, the sum score of the EF rating questionnaire was included as the first block of predictor variables in the regression model and all three EF tasks were jointly included as a second block. For all domains of academic achievement, except for spelling, the change in the amount of explained variance improved significantly when direct assessment of EF was included into the model (p<.05). This trend was also evident for spelling but did not reach statistical significance (p=.088). The amount of explained variance of each domain of academic achievement, however, remained fairly low (reading performance: R2=.125; mathematical performance: R2=.132; spelling performance: R2=.144). In conclusion, the results suggest that direct assessment and kindergarten teacher rating of EF contribute distinctly to the prediction of academic achievement. Including direct assessment of EF in addition to teacher ratings significantly improves the prediction of academic achievement at the end of the first year of primary school. These findings highlight the added value of direct assessment of EF beyond kindergarten teachers’ behavior observations of EF for the prediction of academic achievement.

Authors

• Katharina Thieken, University of Hildesheim
  Presenting Author

• Claudia Maehler, University of Hildesheim
  Non-Presenting Author