Cognitive, Noncognitive, and Home Environment Correlates of Reading Difficulties in Primary-Grade Students with Language Impairment

Purpose: This article first aimed to examine the cognitive (rapid automatized naming, phonological awareness, working memory, nonverbal cognition, and language) correlates of reading difficulty in children with language impairment (LI). Second, we considered whether noncognitive (effortful control, social competence, and behavior problems) correlates offered any explanatory value above that of cognitive factors. Third, we examined whether home environment (specifically household organization and home learning environment) would offer an additional explanatory value. Method: The sample included 165 children in kindergarten and Grade 1 who were receiving intervention for LI in public schools. Standardized measures along with parent interviews/questionnaires were administered at the end of the school year. Results: Logistic regression models indicated the noncognitive factors added discriminatory value to that of cognitive factors in predicting reading difficulties, whereas household factors did not. In the final model using all 11 predictors, prediction accuracy was 88.7% for the typical reading group and 54.2% for the reading difficulty group, with an overall accuracy of 76.4%. Only phonological awareness and working memory significantly contributed to predicting reading group membership when measured in kindergarten and Grade 1. Conclusions: For this sample of children with LI, the most important predictors of reading were cognitive. The child's behavior and social competence improved prediction to a limited but statistically significant degree, whereas home environment did not. Overall classification was low, as only half of the children with reading difficulties were correctly predicted. Important factors differentiating good and poor emergent readers with LI were not captured in this study.