Using computational modelling to understand developmental language disorder (DLD)

In this work package we addressed the underlying causes of developmental language disorder (DLD).

While much is known about the specific difficulties faced by children with DLD, we do not know how these problems arise. Computational modelling is a way to develop potential explanations for the mechanisms underlying typical and atypical development.

We constructed computational models that simulated word learning in typically developing children. Then, we changed different aspects of the function of the model and observed the types of word learning deficits that arose. We compared these ‘atypical’ models with the data on word learning in children with DLD.

Doing this enabled us to specify what might be different in the mechanisms of word learning in children with DLD compared with typically developing children, and it might suggest interventions that can then also be tested with these models.

Project Team: Sam Jones, Bob McMurray, Katie Twomey, Gert Westermann (Lead)

Duration: 3 years, starting 1 May 2020

Project number: 1.1

Key Outputs

Jones, S. D., Stewart, H. J., & Westermann, G. (2024). A maturational frequency discrimination deficit may explain developmental language disorder. Psychological Review, 131(3), 695–715.

Jones, S. D., Jones, M. W., Koldewyn, K., & Westermann, G. (2024). Rational inattention: A new theory of neurodivergent information seeking. Developmental science, pp. e13492

Jones, S. D. & Westermann, G. (2022). Under-resourced or overloaded? Rethinking working memory deficits in developmental language disorder. .Psychological Review, 129(6), 1358–1372.

Jones, S. D., & Westermann, G. (2022). Prediction Cannot Be Directly Trained: An Extension to Jones and Westermann (2021) Journal of Speech, Language, and Hearing Research, 65(10), 3930-3933.

Jones, S. D., & Westermann, G. (2021). Predictive Processing and Developmental Language Disorder Journal of Speech, Language, and Hearing Research, 64(1), 181–185