Children with developmental language disorder (DLD) exhibit significant, persistent impairments of language ability relative to children with typical language development (TLD), including deficits in sentence comprehension and word learning. In typical samples, prediction (the ability to use linguistic and nonverbal information to incrementally anticipate upcoming input) facilitates language comprehension, whereas prediction error (the deviation between the predicted and actual input) supports word learning. However, little research exists on prediction processes in individuals with DLD. The long-term goal of the research project is to examine, in typical and atypical populations, the developmental trajectory of prediction as it relates to language acquisition. To this end, the current objective is to compare prediction, prediction errors, and the role of prediction during word learning in 4- to 6-year-old children with TLD and children with DLD using eye tracking in a visual world paradigm. The central hypothesis is that both groups will predict efficiently, but that children with DLD will be less efficient in recovering from prediction errors, which will negatively impact comprehension and word learning.
Aim 1 is to identify differences in the time course of prediction and revision in children with TLD and DLD. It is expected that both groups will efficiently predict target words but that children with DLD will revise from error more slowly than TLD children, suggesting reduced processing efficiency.
Aim 2 is to ascertain the role of prediction mechanisms in word learning between groups. It is expected that prediction will benefit word learning in both groups whereas prediction error will benefit word learning in children with TLD but not children with DLD. Working memory, speed of processing, interference control, and vocabulary size will be assessed as factors driving group differences in prediction and word learning with the expectation that individual differences in these cognitive abilities will be associated with performance in both groups. The expected outcome of the proposed research is the delineation of weaknesses in prediction processes in DLD that may contribute to problems of language comprehension in the moment and problems in building a language system over time. Results will also reveal relations between prediction and other domain-general factors important to language development. These outcomes will lay the foundation for future research on identification methods and interventions tailored to specific patterns of deficits. If funded, the project will assist the applicant in achieving her goal to have an established program of research on developmental trajectories of language knowledge, language processing, and domain-general cognitive abilities in typical and disordered language. The specific training goals of this project are to 1) extend experience in and knowledge of language disorders, 2) gain experience with eye tracking as a novel behavioral method, 3) develop knowledge of advanced statistical analysis, and 4) develop independence as a researcher. The research and mentoring experience of the sponsor along with the extensive resources of the sponsoring institution will contribute to the success of the project and completion of these goals.
Developmental language disorder (DLD) affects approximately 7% of the population and is defined by significant, persistent language problems including deficits in sentence comprehension and word learning: it is a significant public health concern, yet it remains understudied relative to other neurodevelopmental disorders and its underlying mechanisms are poorly understood. Evidence for the importance of prediction processes in comprehension and learning in typical samples combined with evidence of interference deficits associated with DLD suggest that interference-related prediction deficits may underlie comprehension and world learning deficits in this population. Results from the proposed project will contribute to the knowledge base of mechanisms underlying DLD and will lay the foundation for future research on more targeted and effective identification and intervention approaches.
