Concordance between teacher/therapist opinion and formal language assessment scores in children with language impairment.

As part of a wider study, 242 children attending language units attached to mainstream schools were assessed on a number of formal language assessments. In addition, each child's teacher was asked to state whether, for each of four areas of language difficulty, the participant exhibited this impairment. The four areas were: articulation, phonology, syntax/morphology and semantic/pragmatic impairment. This report compares test results between children thought by their teacher or therapist to show each impairment and those not thought clinically to show the difficulty. Investigations into suitable cut-off scores for the tests used were also examined using level of agreement between the two modes of assessment as criteria. For articulation, phonology and syntax/morphology, teacher opinion was found to discriminate significantly on at least some formal tests. However, for children with semantic/pragmatic impairment, no tests used identified their problems. Furthermore, by use of 25th centile cut-off scores, 66% agreement levels between teacher and test groupings were found in all but the group with semantic/pragmatic impairments. The importance and limitations of analysis of teacher-test concordance is discussed as well as the current lack of suitable formal assessment material for semantic and pragmatic language impairment.