Poster #3 - Middle Childhood Executive Functions and Parenting Behaviors Predict Adolescent Academic Skills

• In Event: Poster Session 19

Abstract

Wide-ranging factors contribute to early academic competence, including executive functions (EF; Blair & Razza, 2007) and parenting behaviors (Englund et al., 2004). When assessed at age 4, parent scaffolding behaviors positively predict, parent negative behaviors negatively predict, and various child executive functions positively predict math and reading academic performance a year later (Devine et al., 2016). These are consistent research findings for early childhood, but there is lack of consistency in the research literature for older children. Wilson and colleagues (2021) reported that an EF composite predicted word reading but not numerical operations in a group of 5- to 12-year-olds. In contrast, Ahmed and colleagues (2019) noted that working memory at age 4 predicted different aspects of academic achievement at 15. Most of these studies failed to control for child IQ and typically only studies of preschool children include parenting contributions to academic skills. We used a longitudinal design to examine EF and parenting behaviors during middle childhood and academic skills in adolescence, controlling for verbal IQ.

Participants were 78 children who contributed data at ages 9 and 14. At the age 9 visit, children did a battery of EF tasks, including backwards digits span (working memory; variable = span), Wisconsin Card Sort Task (shifting/cognitive flexibility, variable = conceptual level), and number Stroop (inhibitory control; variable = mixed condition reaction time). They also completed the Peabody Picture Vocabulary Test (PPVT; variable = standard score) as an assessment of verbal IQ (covariate). At the age 9 visit mothers completed the CCNES (Coping with Children’s Negative Emotions Scale; Fabes et al., 2002; variable = non-supportive parenting scale). At the age 14 visit, adolescents completed the Woodcock Johnson IV (WJ) Applied Problems (math word problems) and Passage Comprehension (reading and understanding written text). Variables of interest were the standardized scores for math and reading. Descriptive statistics are shown in Table 1.

Two different hierarchical regressions were used to predict WJ Applied Problems and WJ Passage Comprehension from PPVT (Step 1 covariate) and the three executive function tasks (Step 2) and non-supportive parenting (Step 3). For Applied Problems, all three steps of the model were significant, accounting for 46% of the variance in math performance. As seen in Table 2, significant predictors in Step 3 included covariate PPVT (beta=.45,p<.001), Stroop reaction time (beta= -.27,p=.007), backwards digit span (beta=.21,p=.03, and non-supportive parenting (beta= -.20,p=.04). For Passage Comprehension, the first two steps of the model were significant, accounting for 43% of the variance in the reading assessment. As seen in Table 2, significant predictors in Step 2 included covariate PPVT (beta=.54,p<.001) and WCST (beta=.23,p=.02).

Results suggest different pathways to math and reading performance on standardized assessments in adolescence. After controlling for verbal IQ, Applied Problems performance was predicted by middle childhood inhibitory control, working memory, and parenting. On the other hand, after controlling for verbal IQ, Passage Comprehension performance was predicted by middle childhood shifting/cognitive flexibility. Future research on adolescent academic skills should focus on individual EF indicators, rather than EF composites, as well as parenting behaviors.

Authors

• Vanessa Paola Díaz Benítez, Virginia Polytechnic Institute and State University
  Presenting Author

• Martha Ann Bell, Virginia Polytechnic Institute and State University