Enhancing English Pronunciation with Windows Speech Recognition Training: A Preliminary Study

Abstract

Teacher educators have developed great interest in applying automatic speech recognition software to improve English learners’ pronunciation. However, these studies are few compared with those emphasizing on other skills such as reading and writing. This case study thus aims to fill this gap by adopting the Windows Speech Recognition (WSR) system to help English learners improve pronunciation skills and promote learner autonomy. Drawing on the Interaction Hypothesis in Second Language Acquisition, this study sought to answer two questions: (1) with the use of WSR in pronunciation training, what is learner performance of pronouncing /n/ and /l/ sounds? and (2) what are learner attitudes toward this training? The training lasted for three weeks, with one hour each time and twice a week. Data sets included a pre-test (screening test) and a post-test based on screen- and audio-recordings of training sessions, a questionnaire, and a semi-structured interview with students. Field notes were also recorded when the entire training process was observed. Findings revealed that learner performance shown in the screening- and posttest is not significant. However, it indicates that with appropriate training, it is possible for learners to understand the feedback provided by WSR and apply the knowledge to judge their own practices. In addition, since learners held positive attitudes towards the use of this software, it is suggested that teachers or tutors can integrate the software with their curriculum for improving English students’ pronunciation skills of segmental features, their learner autonomy and learning strategy.