# Prediction and Word Learning in Young Children with Autism Spectrum Disorder > **NIH NIH F31** · UNIVERSITY OF WISCONSIN-MADISON · 2024 · $41,290 ## Abstract ABSTRACT Current estimates suggest that autism spectrum disorder (ASD) impacts 1 in 44 children in the United States. Many young children with ASD demonstrate language delays, which are often the first area of concern noted by their parents. Young children with ASD show markedly different word learning profiles than their neurotypical (NT) peers, characterized primarily by atypical or delayed use of word learning mechanisms. We currently have an incomplete understanding of these differences. Discerning the mechanisms that underlie language learning deficits in ASD will afford greater success in alleviating these difficulties and supporting language learning in this population. One theoretical account, a predictive impairment in autism (PIA) hypothesis, has shown promise for explaining some phenotypic characteristics of ASD (i.e., difficulty making social predictions). The utility of this theoretical framework for explaining difficulty with language learning, however, remains largely unclear. Autistic children show intact statistical word learning abilities on experimental tasks. However, given that children learn words in their natural, often unpredictable environments, difficulty tracking unpredictable stimuli might have profound impacts on autistic children’s word learning. The overall objective of the proposed project is to determine how established, domain-general difficulties with prediction in ASD might impact novel word learning. This project will employ established eyegaze methodology to test the impact of predictable and unpredictable presentations on word learning in this population. The central hypothesis is that unpredictable contexts during training will disproportionately disrupt word learning in children with ASD and result in disproportionate detriments to the retention and generalization of object-label pairings. Preliminary data suggest that autistic children are disproportionately disrupted by unpredictable input in a fast-mapping task. Guided by the PIA hypothesis and this strong preliminary data, the proposed project will address three specific aims. Specific Aim I will characterize the impact of predictable and unpredictable presentations of novel-word object pairings on initial word learning (fast-mapping) in autistic toddlers compared to cognitively matched NT peers. Specific Aim II will investigate the impact of predictability during initial learning on autistic toddlers’ retention and generalization of word-object pairings over a short (5-minute) delay, compared to NT peers. Specific Aim III will evaluate the extent to which child characteristics (e.g., language and cognitive abilities) predict individual variability in word learning performance following predictable and unpredictable input in autistic children. The proposed project will be the first to apply established eyegaze methodology to test this theoretical framework and its application to initial word learning, retention, and generalization in this population. T... ## Key facts - **NIH application ID:** 10739296 - **Project number:** 5F31DC020902-02 - **Recipient organization:** UNIVERSITY OF WISCONSIN-MADISON - **Principal Investigator:** JANINE R MATHEE-SCOTT - **Activity code:** F31 (R01, R21, SBIR, etc.) - **Funding institute:** NIH - **Fiscal year:** 2024 - **Award amount:** $41,290 - **Award type:** 5 - **Project period:** 2022-09-01 → 2024-07-19 ## Primary source NIH RePORTER: https://reporter.nih.gov/project-details/10739296 ## Citation > US National Institutes of Health, RePORTER application 10739296, Prediction and Word Learning in Young Children with Autism Spectrum Disorder (5F31DC020902-02). Retrieved via AI Analytics 2026-05-26 from https://api.ai-analytics.org/grant/nih/10739296. Licensed CC0. --- *[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*