Participant evaluations of rate and communication efficacy of an AAC application using natural language processing

This research explores the efficacy of an AAC application, Converser, that uses natural language processing to assist in communication. Converser uses speech recognition of the speaking partner to predict contextually appropriate utterances. This article reports on the subjective data gathered from an experimental evaluation of Converser's efficacy (see Wisenburn and Higginbotham, 2008 for a full description of Converser and the objective results). Converser was evaluated in two communication tasks (a Conversation and an Interview task) under two conditions: a simple alphabet board without Converser (alpha-only condition), and an identical board with Converser (alpha-Converser condition). Subjective data was gathered through rating questionnaires and written comments. Program users rated the speed of communication faster in the alpha-Converser condition. Program user ratings of quality, and speaking partner ratings of speed and quality, showed no difference between the two conditions; however, the participant comments about Converser were positive.