Abstract

Low language proficiency remains a significant barrier to healthcare access for many patients throughout the world. Training healthcare professionals in the language of their minority language patients, therefore, should lead to greater healthcare access, lowered costs, better health outcomes, and improved patient satisfaction (Zambrana, Molnar, Munoz, & Lopez, 2004). One important aspect of language training involves the development of accessible, appropriate, and pedagogically sound language training materials. The first goal of this thesis is to describe the development of the "Virtual Language Patient," a computer-based language training module based on the Virtual Dialogue Method (Harless, Zier, & Duncan, 1999). The prototype system under consideration employs automatic speech recognition (ASR) technology, using video clips of a simulated medical history interview with a minority language patient. The second goal of this thesis is to report the findings of a proof-of-concept feasibility study where the ease of operability and fitness of purpose of this prototype system were explored. Five nursing-students at a French language nursing college in Quebec reported the system to be easy to operate and fit for their anticipated language learning needs in terms of target language, choice of interlocutor, mode of interaction, task type, and corrective feedback. Training effects on participants' pronunciation scores, speech rate, and sense of preparedness for real life medical interviews suggest that the system can be effective in language training for healthcare professionals. Implications for the improvement of this and future virtual dialogue systems are also discussed.