Purpose: This study examines the classification accuracy and interrater reliability of a dynamic assessment (DA) of inferential word learning designed to accurately identify kindergarten through sixth-grade students with and without language disorder. Method: The participants included 127 school-age children from a mountain west school district who were administered a DA of inferential word learning that entailed a pretest, a teaching phase, an examiner rating of the child's ability to infer word meaning (modifiability), and posttests. Results: Hierarchical logistic regression and receiver operator characteristic (ROC) analyses revealed that combining all posttests, the modifiability total, and the final examiner judgement scores from this DA yielded the strongest sensitivity (.83) and specificity (.80). The static measures and the dichotomized final examiner judgement had excellent reliability; yet the individual modifiability measures (with the exception of disruption and frustration) had poor reliability. Conclusion: In concordance with a previous study, results indicate that a dynamic assessment of inferential word learning may be an efficacious method of identifying language disorders in school-age populations.

Purpose: This study examines the classification accuracy and interrater reliability of a dynamic assessment (DA) of inferential word learning designed to accurately identify kindergarten through sixth-grade students with and without language disorder. Method: The participants included 127 school-age children from a mountain west school district who were administered a DA of inferential word learning that entailed a pretest, a teaching phase, an examiner rating of the child's ability to infer word meaning (modifiability), and posttests. Results: Hierarchical logistic regression and receiver operator characteristic (ROC) analyses revealed that combining all posttests, the modifiability total, and the final examiner judgement scores from this DA yielded the strongest sensitivity (.83) and specificity (.80). The static measures and the dichotomized final examiner judgement had excellent reliability; yet the individual modifiability measures (with the exception of disruption and frustration) had poor reliability. Conclusion: In concordance with a previous study, results indicate that a dynamic assessment of inferential word learning may be an efficacious method of identifying language disorders in school-age populations.