Similarities in lexical predictive processing in children with and without dyslexia
Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Evidence is mixed regarding whether or not individuals with dyslexia struggle with higher-level spoken language comprehension, including predictive processing. We posit that this uncertainty may be explained by experiments’ task designs—with decontextualized tasks suggesting more extensive differences in dyslexia and more naturalistic tasks suggesting equivalent language comprehension in dyslexia. Children with dyslexia (n=73) and typical readers (n=96) listened to a story during fNIRS neuroimaging. Focusing on predictive processing as a measure of higher-level language comprehension, we used GPT-2 to quantify each word’s predictability within the story’s context, allowing us to identify brain regions engaged in predictive processing. Our findings indicate similar neural responses to word predictability for children with dyslexia and typically developing readers during naturalistic listening. These results highlight the use of contextual lexical prediction during naturalistic language processing as a potential strength in children with dyslexia, while underscoring the continued importance of dyslexia screening and intervention.
Source: PsyArXiv (OSF Preprints)
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7204087030
