Many children have developmental language disorder, which makes it difficulty to learn language, including vocabulary, and results in ongoing academic and social difficulties. Despite the fact that most words are learned in context without direct teaching, we know very little about how children with developmental language disorder learn words in context. This project will combine, for the first time, two strategies which improve contextual word learning in children with typical development, and test their effect in both typically developing children, and those with developmental language disorder. The results will provide timely information that will contribute to evidence based practice for contextual word learning in children with developmental language disorder.

Vocabulary achievement impacts important life outcomes such as academic success, vocational options, and earning potential. Developmental language disorder is a common disorder which impacts word learning, both when words are directly taught and when words are learned implicitly after multiple contextual encounters. Contextual word learning is the most common situation in which words are learned, and this typically occurs while reading. As such, interventions which increase the effectiveness of contextual word learning have the potential to meaningfully alter the trajectory of vocabulary growth.

The long term objective of this research program is to develop effective interventions for children with developmental language disorder, which target contextual word learning while reading. The overall aim of this study is to test the effectiveness of an intervention which combines two approaches known to be effective in typical development: metalinguistic training about semantic inferencing, and use of multiple texts with high semantic diversity. Semantic diversity refers to the range of semantic content in texts which contain the novel words. In Aim 1, we test the effectiveness of this novel intervention in children with typical development (n=50), and in Aim 2, we evaluate its effect in children with developmental language disorder (n=50). Participants will complete a three session metalinguistic intervention involving repeated meaning generation, explanation about proposed meanings, and feedback about accuracy. The quality of semantic inferencing about novel words which are untreated (not involved in the intervention) will be measured pre- and post-intervention as a measure of the effect of intervention on the skill of semantic inferencing. Measures of retention of the semantics of words used in intervention will be taken one week after each session, and are expected to increase with more exposure to the metalinguistic intervention. Within sessions, words will be presented in three different semantic diversity contexts: no diversity (repeated stories), medium diversity (unique stories with the same global context), and high diversity (unique stories with different global contexts). Based on prior research about the effect of semantic diversity on word learning in children and adults, we predict that measures of retention of the semantics of words used in intervention are expected to be highest in the high semantic diversity condition.

Results from this study will be used to support an application for a fully powered randomized controlled trial of a contextual word learning intervention for children with DLD, and will inform best practices for children with typical development. The project is innovative because we will examine contextual word learning in developmental language disorders, we will combine two treatment approaches in a novel way, and because we focus not only on the effect on treated words, but also on generalization of the skill of semantic inferencing.

Inclusion Criteria:

• Aim 1: Standard Score > 90 on Clinical Evaluation of Language Fundamentals - Core Test Language Battery administered in 2nd grade
• Aim 2: Standard Score <= 85 on Clinical Evaluation of Language Fundamentals - Core Test Language Battery administered in 2nd grade

Exclusion Criteria:

• hearing impairment
• uncorrected visual impairment

Deidentified data will be made available on Open Science Foundation.

Data will be made available upon completion of data collection and coding, or six months after the funding period is complete, whichever comes first. Data will not be removed from public access.

Open Science Framework is a public access site.