[Diagnostic Accuracy of the LiSe-DaZ for Children with Specific Language Impairment]

Objective: Currently, only few tests for the development of speech and language exist for bi- or multilingual children in Germany. One of those, the LiSe-DaZ (Linguistic performance measurement - German as a second language), was examined in a prospective study regarding its practicability and the sensitivity to detect children with specific language impairment in a group of children aged 5 to 7 who suffered from a severe language impairment according to clinical tests.

Material and methods: 74 children (mean age: 60 months; 46% monolingual German-speaking; 54% bi- or multilingual) with severe specific language impairment were examined with the LiSe-DaZ in addition to the clinical established diagnostic during their in-patient stay in the hospital.

Results: The children, on average, showed in the receptive language abilities (LiSe-DaZ vs. TROG-D), the expressive vocabulary (LiSe-DaZ vs. AWST-R or WWT) and in the use of prepositions (LiSe-DaZ vs. Ravensburger Dysgrammatical clinical trial) significantly (p<0,0005) better results in the LiSe-DaZ. Thus, the majority of children were diagnosed as language impaired by clinically established tests whereas the LiSe-DaZ considered the children's language development to be normal. This difference was consistently more prominent for children with German as a second language.

Conclusions: Compared with the clinically established tests, the informative value of the LiSe-DaZ turned out to be insufficient. The LiSe-DaZ does not detect children with the need of language therapy. Nevertheless, a norming of the established speech tests for bi- or multilingual children would be desirable to avoid unfounded judgements.