Using a Database of Typical Speakers to Describe the Expository Language Skills of Children with Language Impairment

Abstract

Purpose. This study examined the expository language skills of older students with language impairment (LI) in relation to a large database of typically developing (TD) students. The purpose of the study was to investigate whether comparing language samples to the database allowed users to distinguish between adolescents with LI and those with typical language, and develop individual profiles of relative strengths and weaknesses in children with LI. Methods. School speech-language pathologists elicited expository language samples from high school students with LI (N = 9; mean age = 16;8 [years;months]) by asking them to explain how to play their favorite game or sport as if speaking to a naïve listener. Language samples were transcribed using Systematic Analysis of Language Transcripts (SALT; Miller & Iglesias, 2012), and analyses were completed to compare the performance of students with LI to a large database of samples from TD children in nine language measures. To develop individual expository language skill profiles of relative strengths and weaknesses, the scores of students with LI were compared to their own overall performance on the expository task. Individual profiles were compared to determine whether subgroups of LI appeared. Results. Analysis revealed distinct profiles of relative strengths and weaknesses for eight of the nine participants. One student with LI demonstrated a relatively equal level of performance across all language measures. When individual participants’ scores were compared to the database of TD peers matched in chronological age, all adolescents with LI demonstrated performance at least one standard deviation lower than the database mean in at least two language measures. Conclusions. Expository language sample analysis facilitated the development of individual profiles of strengths and weaknesses in this sample of adolescents with LI. Analysis of expository performance in a larger sample of older students with LI will help determine the number and compositon of linguistic profiles, which specific language measures are most effective in differential diagnosis of LI, and whether this expository task is effective in distinguishing students with LI from those with TD language. Due to the small sample size, the results of this study should be considered preliminary and interpreted with caution.