Jennifer Windsor - US grants

There is a large population of children with language impairments (LI) who are at risk for academic and social failure. While children with LI show most difficulty with grammar, they also show limitations in other aspects of language and show subtle cognitive and perceptual-motor problems. To explain this breadth of difficulties, current research has focused on the idea that children with LI have more limited information-processing capacities than other children of the same chronological age. Because they have more limited capacities, the task performance of children with LI is more vulnerable to competing resource demands. When resources are strained, task performance is impaired. This idea is intuitively appealing because of its possible explanatory power and has received widespread attention. However, it has received little systematic examination. Moreover, the methods used to explore limited processing accounts have not been evaluated critically. The long-term goal is to clarify the processing skills of children with LI. The aim of the proposed project is to test a central hypothesis of limited processing capacity accounts, the generalized slowing hypothesis. In its strongest form, the hypothesis states that children with LI are proportionally slower to complete language, cognitive, and perceptual-motor tasks compared to age-matched children. Regardless of specific task characteristics, the processing speed of children with LI should increase linearly across tasks as a simple multiplicative function of the processing speed of age-matched peers. The hypothesis is tested in three studies that provide converging evidence about the hypothesis and indicate predictors of relative processing speed for children with LI. One hundred children with and without LI participate in 32 experimental tasks that vary task complexity systematically. It is anticipated that generalized slowing will be apparent for children with LI, but that there will be different slowing rates in the language, cognitive, and perceptual-motor domains. Project results will inform theoretical approaches to LI and thus facilitate more effective assessment and intervention strategies for children with LI.

[unreadable] DESCRIPTION (provided by applicant): The goal of this exploratory project is to develop and refine a novel battery of processing-based measurement tools that can be used to effectively identify a behavioral phenotype for primary or "specific" language impairment (LI) across children with diverse language learning experiences. Participants will be English-only speakers with and without LI, bilingual speakers of Spanish and English, and bilingual speakers of Hmong and English. We propose 2 types of tasks that emphasize either language processing or the processing of nonlinguistic information. The 3 language-based processing tasks are non-word repetition, counting span and rapid naming, in English, Spanish and Hmong. The 3 non-linguistic processing tasks are choice visual detection, auditory serial memory, and Tallal repetition. Group comparisons followed by likelihood ratios will be used to calculate task sensitivity and specificity. Logistic regression and cluster analyses will be used to determine the best combination of tasks for identifying children with LI and potential LI subgroups. The innovation of this research application is 2-fold: to take the fundamental step of addressing the LI phenotype within the context of linguistic diversity and to develop assessment measures that tap into robust aspects of the information processing system, that is, to examine the nonlinguistic cognitive profile accompanying LI language performance as well as the more observable deficit in language processing. Results from the proposed project will directly contribute to a more clearly defined LI phenotype as well as to the development of non-biased assessment measures needed for separating typical language development from LI in an increasingly diverse population. [unreadable] [unreadable]