"Testing longitudinal associations between executive function and academic achievement": Correction to Willoughby, Wylie, and Little (2019).

Abstract

Reports an error in "Testing longitudinal associations between executive function and academic achievement" by Michael T. Willoughby, Amanda C. Wylie and Michael H. Little (Developmental Psychology, 2019[Apr], Vol 55[4], 767-779). In the article, the authors report a coding error. Specifically, in the ECLS-K dataset, there are two variables that refer to free/reduced-price lunch. An incorrect variable was used to construct the free/reduced price lunch subsample, which resulted in the over-selection of students into the free/reduced-price lunch subsample (i.e., 2,830 were included but only 1,910 should have been). The authors reanalyzed all data using the corrected free/reduced price lunch subsample. While all the numerical values related to the free/reduced-price lunch subsample that were reported in the original text and tables changed, the magnitude of changes were trivial and none of the substantive conclusions made in the article changed. Updated results are available from the first author. (The following abstract of the original article appeared in record 2018-65595-001.) Children with higher levels of executive function (EF) skills consistently demonstrate higher levels of academic achievement. Despite the consistency of these associations, fundamental questions remain about whether efforts to improve an individual child's EF skills result in corresponding improvements in his or her academic performance. In the absence of experimental evidence, developmentalists have used repeated measures designs to test the nature, magnitude, and direction of the associations between EF skills and academic achievement. In contrast to previous studies, this study described how between- and within-person associations between EF and achievement address different questions. Using data from a subsample of participants (N = 6,040) from the Early Childhood Longitudinal Study-Kindergarten, 2010-2011 (ECLS-K:2011) cohort, we estimated a series of latent growth curve models with structured residuals to test the between and within-person associations between 2 dimensions of EF (working memory, cognitive flexibility) and 2 domains of academic achievement (math, reading). Whereas between-person associations between EF and achievement were large (φ = .55-.91), the within-person associations were small (βs = -.10-.25). Within-person effects of earlier reading achievement on later EF skills was the most consistent finding. Results were unchanged when analyses were repeated using the subset of children who were eligible for free and reduced-price lunch, a proxy for low socioeconomic households. Results are discussed with respect to interest in improving EF skills as a means for facilitating school outcomes. (PsycINFO Database Record (c) 2019 APA, all rights reserved).