Intelligibility Measures of Dysarthric Speech

Ten dysarthric and 10 matched normal speakers recorded three word lists (50 CNC’s, 60 Clusters, and 50 Spondees). Intelligibility scores were obtained for lists, order of presentation, list by order of presentation, and lists and order of presentation combined. Results indicate that intelligibility testing can be employed to evaluate dysarthric speech. It was possible to differentiate among dysarthrics in terms of intelligibility based on listener judgments. The rank of the dysarthrics remained relatively stable across the intelligibility measures. No significant differences between lists or order of presentation were obtained. Significant differences for dysarthrics and the dysarthric by list, and dysarthric by order interactions were obtained. The intelligibility scores for dysarthrics and normals were markedly different.