LSA for diagnosing language disorder in Spanish–English bilingual children

This study looked at language samples from 40 Spanish-English bilingual children—20 with developmental language disorder and 20 with typical language. Samples were elicited in Spanish (L1) using wordless picture books. The children repeated a story read to them (retelling) and generated a story on their own (storytelling). Samples were transcribed with SALT and CLAN and coded for: “…lexical diversity (D)*, grammaticality (grammatical errors per communication unit [GE/CU]), sentence length (mean length of utterance in words [MLUw]), and sentence complexity (subordination index [SI]).” Key findings are in the chart below (click to make it bigger and easier to read):


The results suggest that LSA by itself was fairly accurate at diagnosing language disorder in this population, and that different elicitation methods bring out different skills. However, note that this study included only a small sample of 4- to 5-year-old Spanish-English bilinguals whose language was sampled in Spanish. These results may not apply to other age groups, or to monolingual Spanish- or English-speaking children with language disorder. Overall, this study reiterates that LSA is an important part of a language assessment for bilingual children.

*D is a measure of lexical diversity computed by CLAN that is less influenced by sample size than measures like type token ratio (TTR). Remember, CLAN is free! All other measures were computed by SALT.

Kapantzoglou, M., Fergadiotis, G., & Restrepo, M. A. (2017). Language sample analysis and elicitation technique effects in bilingual children with and without language impairment. Journal of Speech, Language, and Hearing Research. Advance online publication. doi: 10.1044/2017_JSLHR-L-16-0335.