An Evaluation of CAST

Abstract

This study assessed the effectiveness of CAST (Computer-Aided Speechreading Training program). Two groups of 8 normal-hearing adults completed a pretraining visual speech perception test protocol that consisted of a Visual-Consonant Recognition Test, a test of Sentence Understanding Without Context, a test of Sentence Understanding With Context, and a semiautomated modified Continuous Discourse Tracking activity available with CAST. One group completed the CAST program. A posttraining test protocol was administered to the subjects 2 weeks following the maximum time provided to complete the training program (i.e., 10 weeks). Independent t tests revealed no differences between the control group and the experimental group in the percentage improvement score obtained on the following measures: viseme recognition score, Sentence Understanding Without Context (key word and total word recognition scores), and Sentence Understanding With Context (key word recognition score). Significant differences between the groups were found for the total word recognition score on the Sentence Understanding With Context test and the CAST modified Continuous Discourse Tracking activity. The results indicate that CAST was most effective in developing aspects of synthetic visual speech perception skills.