Presenter: Tori Rydalch, McKay School of Education, Communication Disorders
Authors: Tori R., Samantha S,, Brittanie H., Courtney C., Jillian A., Sarah H., Nichole P., Riley B., Kristi Jones
Faculty Advisor: Douglas Petersen, McKay School of Education, Communication Disorders
Institution: Brigham Young University
Many norm-referenced tests (NRTs) lack an adequate degree of sensitivity and specificity (i.e., classification accuracy) for identifying language disorders in school-age children (Spaulding, Plante, & Farinella, 2006), particularly when assessing culturally and linguistically diverse children (Laing & Kamhi, 2003). Dynamic assessment is a promising alternative approach to NRTs. Dynamic assessment has been shown to mitigate bias in identifying school-age children with language disorders because, unlike NRTs, dynamic assessment does not assess a child’s static, current knowledge, rather, dynamic assessment measures a child’s ability to learn and respond to evidence-based instruction. Dynamic assessment often uses a test-teach-retest model (Peña & Iglesias, 1992). Several studies have investigated the potential of dynamic assessments of language to identify language disorders in children (Orellana, Wada, & Gillam, 2019). Although the research evidence is promising, more research is needed to examine the validity of a dynamic assessment of language across a large representative sample of diverse students. Therefore, the purpose of this study was to examine evidence of classification accuracy of a dynamic assessment of narrative language across a large sample of students. Participants included approximately 350 diverse students with and without language disorders from kindergarten through sixth grade from multiple states. Prior to administering the dynamic assessment, language disorder was established if a student had a current IEP for language and performed poorly on a non-word repetition and an oral narrative language sample. The dynamic assessment entailed four distinct steps: pretest, teaching, modifiability rating, and post-test. Examiners were blinded as to whether students had a language disorder or not. Results indicate that sensitivity and specificity were above 80%, with the combination of posttest scores and modifiability scores from the dynamic assessment yielding the greatest classification accuracy. Inter-rater reliability was above 85% for all sections of the dynamic assessment.