Different Dimensions of Language Input Support Language and Executive Functioning Development in Preschool

• In Event: The Role of SES, Stress, and Language Input on Cognitive Development in at-risk Children

Abstract

Childhood socioeconomic status (SES) is correlated with developmental differences across many cognitive domains, and two of the most affected domains are language and executive functioning (Pace et al., 2017; Lawson et al., 2018). Although language and executive functioning (EF) both develop rapidly in the preschool years, a growing body of research suggests that early language development supports subsequent EF development (Bruce & Bell, 2022). Additionally, SES is frequently correlated with various aspects of children’s early language input (Rowe & Snow, 2020), which in turn predicts differences in language development trajectories. However, less is known about how this language input may also influence trajectories of EF development, and whether the same input features predict development across both domains.

A large (n=275, 50% male), longitudinal, SES-diverse sample of children completed repeated assessments of language (sentence comprehension) and EF (cognitive flexibility, behavioral inhibition, and cognitive inhibition) at four waves across ages 3-5 years. At the first wave, mother-child dyads engaged in a 7-minute free-play session, and videos were transcribed verbatim at the utterance level in CLAN using CHAT conventions (MacWhinney, 2000). Six measures were extracted from maternal speech and served as indicators to latent factor representing the three input dimensions proposed by Rowe & Snow (2020): a linguistic factor was determined by syntactic complexity (mean length of utterance) and lexical diversity (VOC-D; Malvern et al., 2004); an interactive factor was determined by lexical contingency (maternal repetition of child-spoken word roots) and the number of wh-questions; and a conceptual factor was determined by the frequency of decontextualized utterances (past, future, scripts) and higher order thinking talk (explanations, connections). Latent growth models were estimated to evaluate bidirectional relationships between language and EF development, the influence of socioeconomic status (maternal education and income-to-needs ratio), and the potential mediating role of language input dimensions, controlling for child gender.

Although language and EF development were bidirectionally related, language comprehension more strongly predicted EF at the earliest ages. SES was positively associated with initial language and EF levels (intercepts), but not rates of growth, indicating early and persistent influences. Language development was predicted by both the linguistic and interactive dimensions, while EF development was predicted by both the conceptual and interactive dimensions, and these input factors mediated SES differences in both cognitive domains. Finally, language measures significantly explained SES differences in EF development, indicating both direct and indirect routes by which language input dimensions influence EF development.

Findings have important conclusions for the co-development of language and EF in preschool, and the role of SES and language input. Results suggest that children’s language input not only predicts language development, but also their EF development. Specifically, contingent interaction supports both cognitive domains, and higher order conceptual language notably supports EF development. Additionally, results suggest that SES differences in EF development are explained, at least in part, by early differences in language development and language input. These findings have implications for early interventions targeting multiple dimensions of children’s language input, to ultimately reduce disparities in language and EF development.

Authors

• Rachel R. Romeo, University of Maryland - College Park
  Presenting Author

• Meredith L. Rowe, Harvard University

• John Coleman Flournoy, Harvard University

• Klaudia Izabela Kulawska, Lehigh University

• Amy E Carolus, University of North Carolina at Chapel Hill

• Grace Kim

• Liliana J Lengua, University of Washington

• Katie McLaughlin, Harvard University