Speech recognition technology will eventually aid speech therapists in the many time-consuming drill-and-practice sessions necessary for effective therapy. The question being asked in this proposal is whether or not the technology is currently sophisticated enough to be of benefit. Resources assembled to answer this question include: (1) the cooperation of the Division of Communication Disorders at North Texas State University, including as co-principal investigator the Coordinator of Audiology; (2) an Apple-based authoring system designed for implementing courseware using voice input and output (the system has already been tested as a toll for foreign language training); and (3) the expertise of a company dedicated to speech recognition technology and applications. The approach is divided into four steps. The first is to identify, observe and record specific therapists and clients engaged in targeted repetitive therapies. The second will be to author brief, specific lessons for use in therapy. The third will be to train the therapists to use the system. The fourth will be to record the client's progress and the therapist's time involved. The subjective question to be answered is how well the client and therapist accept the system. The objective questions are whether or not the therapist can provide the client with comparable or superior service with less time, and which, if any, repetitive tasks can reasonably be accomplished with current technologies.
