A PRELIMINARY EVALUATION OF THE CLINICAL EFFECTIVENESS OF VOWEL TRAINING FOR HEARING-IMPAIRED CHILDREN ON IBMS SPEECH VIEWER

The clinical effectiveness of the ''Vowel Accuracy Skill Building'' module of IBM's SpeechViewer system was tested by comparing the speech performance of four profoundly hearing-impaired children trained on this system with four profoundly hearing-impaired children who received traditional speech training without computer support. Progress in vowel articulation was continuously measured for both groups using SpeechViewer. Two measures were taken of vowel improvement: the degree of differentiation between the first (F1) and second (F2) formant frequencies for the Absolute value of i vowel and the amount of variation about the second formant frequency value for this same vowel. Although neither measure revealed statistically significant differences between the subject groups pre- to post-training, there was a somewhat better average F2 - F1 differentiation at the end of training for the subjects who used SpeechViewer than for those traditionally-trained. Neither method of speech training resulted in improved variation for F2. The results are discussed within the context of current research and the clinical effectiveness of computer-assisted speech training. Finally, some comments are offered for improving SpeechViewer.