Varying Dimensions of Language Input Predict Language and Executive Function Development Across Socioeconomically Diverse Families

Authors:
Preprint
Publisher website
Google Scholar
Abstract
Children’s early language experiences are linked to their language and executive functioning (EF) development, and vary by socioeconomic status (SES). However, unique features may support different cognitive domains because input is multifaceted. The present study characterizes the dimensional structure of maternal language input and examines how these dimensions vary by SES and predict language and EF development across the preschool years. Thirteen metrics of maternal speech were extracted and/or manually coded from semi-structured play with 275 mother-child dyads when children were 3 years (137 male, 138 female; 65% White, 4% Black, 2% Latino, 2% Asian, 28% multiracial). Children’s language and EF were assessed four times between ages 3 and 5 years (data collected 2008–2012). Exploratory factor analysis revealed four input dimensions: linguistic quantity, linguistic complexity, interactiveness, and conceptual input. These differentially predicted children’s baseline language and EF performance, but not growth. Specifically, interactive and conceptual input predicted language skills, while complexity predicted EF skills. The four dimensions jointly mediated the relationship between SES and baseline language (β=0.053), and a serial pathway linked SES through input and language to EF (β=0.052). These findings reveal distinct pathways through which specific features of early input support children’s language and EF.