Presentation Type
Oral Presentation
Category
Social Sciences/Humanities
Abstract/Artist Statement
Children with developmental language disorder (DLD) have difficulty understanding and using spoken language, which negatively impacts their academic outcomes and psychosocial wellbeing. DLD is highly prevalent, affecting approximately two out of every 25 children, but less than 40% of children with DLD are ever identified to receive services. Currently, we rely on parent and teacher referrals to find DLD, which is clearly insufficient. Additionally, 50% of children with DLD will also develop the reading disorder dyslexia, meaning they will struggle with both spoken and written language. Given schools’ limited time and resources, we need an efficient method for (a) identifying DLD and (b) identifying which children with DLD will go on to develop dyslexia. However, there is currently no method of predicting later language/literacy abilities for children with DLD. This study was the first to investigate three metalinguistic skills in children with DLD—phonological awareness, orthographic knowledge, and morphological awareness—which may longitudinally predict language and literacy abilities. As part of a wider NIH grant, this study followed several cohorts of children in Montana and Massachusetts from kindergarten to second grade. Three subgroups of each cohort—those with DLD only, DLD plus dyslexia, and a control group of typically developing children—completed a battery of assessments in kindergarten and second grade. Multivariate multiple regression will be used to evaluate whether the three metalinguistic skills can (a) concurrently predict kindergarten language and literacy scores (i.e., identify DLD) and (b) longitudinally predict second grade scores (e.g., predict dyslexia). To determine if predictive power varies by subgroup, the regression models will include an interaction term dummy coded for group status. If these three metalinguistic skills predict DLD and/or dyslexia, this study will provide a cost-effective identification method that eliminates the subjectivity of parent/teacher referrals, ultimately helping more children get the services they need.
Mentor Name
Catherine Off
Melissa Phelan video presentation
Predictive Power of Metalinguistic Skills in Children with Developmental Language Disorder
UC 332
Children with developmental language disorder (DLD) have difficulty understanding and using spoken language, which negatively impacts their academic outcomes and psychosocial wellbeing. DLD is highly prevalent, affecting approximately two out of every 25 children, but less than 40% of children with DLD are ever identified to receive services. Currently, we rely on parent and teacher referrals to find DLD, which is clearly insufficient. Additionally, 50% of children with DLD will also develop the reading disorder dyslexia, meaning they will struggle with both spoken and written language. Given schools’ limited time and resources, we need an efficient method for (a) identifying DLD and (b) identifying which children with DLD will go on to develop dyslexia. However, there is currently no method of predicting later language/literacy abilities for children with DLD. This study was the first to investigate three metalinguistic skills in children with DLD—phonological awareness, orthographic knowledge, and morphological awareness—which may longitudinally predict language and literacy abilities. As part of a wider NIH grant, this study followed several cohorts of children in Montana and Massachusetts from kindergarten to second grade. Three subgroups of each cohort—those with DLD only, DLD plus dyslexia, and a control group of typically developing children—completed a battery of assessments in kindergarten and second grade. Multivariate multiple regression will be used to evaluate whether the three metalinguistic skills can (a) concurrently predict kindergarten language and literacy scores (i.e., identify DLD) and (b) longitudinally predict second grade scores (e.g., predict dyslexia). To determine if predictive power varies by subgroup, the regression models will include an interaction term dummy coded for group status. If these three metalinguistic skills predict DLD and/or dyslexia, this study will provide a cost-effective identification method that eliminates the subjectivity of parent/teacher referrals, ultimately helping more children get the services they need.