DEVELOPMENT AND VALIDATION OF A COMPUTER-BASED APHASIA THERAPY FEATURING AUTOMATIC SPEECH RECOGNITION

Abstract

This dissertation is part of a broader project whose mission is to develop a computer-based aphasia therapy (CBAT) that improves currently available applications, by focusing on usability and — critically — feedback based on automated speech recognition (ASR). The research presented here focused on the initial user-centered design of a CBAT prototype, usability testing, and preliminary testing of the efficacy of the ASR-based feedback. To this end, this dissertation set out to: conduct a needs-finding analysis to identify the most common barriers and facilitators in the use of technologies that can inform the design of a prototype; develop and test the usability of a prototype of home-based therapy, featuring an ASR engine, geared for treatment of word-finding skills; conduct a preliminary test of the efficacy of delivering feedback via an ASR. This dissertation provides the basis upon which to integrate additional improvements.