Presentation

Using ASR to Improve Taiwanese EFL Learners’ Pronunciation: Learning Outcomes and Learners’ Perceptions

---

Abstract

Previous research has shown positive effects of Automatic Speech Recognition (ASR) on the development of second language (L2) pronunciation (Liakin et al., 2017) and on learners’ self-perception of their intelligibility (Mroz, 2018). However, empirical studies on the benefits of ASR for pronunciation improvement and learners’ engagement remain scant. The present study therefore investigates the effect of ASR on segmental development and explores learners’ perception of ASR as a learning tool. A total of 47 Taiwanese university EFL students participated in six sessions of autonomous ASR practice over three weeks; for each practice session, participants used the ASR programs on their mobile phones to record a short reading passage and minimal pairs. Vowel duration and F1 and F2 formant frequency values for minimal pair tokens produced at pretest and posttest were analyzed. The results of the acoustic analysis revealed that learners significantly improved the /æ/-/ɛ/ distinction in production, while continuing to struggle with the /i/-/ɪ/ contrast.The participants also reported less positive experience of and attitude toward the use of ASR for pronunciation practice than observed in previous studies. Possible reasons for the findings are discussed.

How to Cite:

Chen, W., Inceoglu, S. & Lim, H., (2019) “Using ASR to Improve Taiwanese EFL Learners’ Pronunciation: Learning Outcomes and Learners’ Perceptions”, Pronunciation in Second Language Learning and Teaching Proceedings 11(1).