EFL Students' Perception of the use of Text-to-Speech Synthesis in Pronunciation Learning

摘要

English as a Foreign Language (EFL) learners usually find speaking particularly challenging. One way to improve speaking skills in a foreign language is to speak with native speakers of the target language as often as possible. However this is not always easy in places where the population of speakers of the target language is scarce. Language anxiety is another major hurdle that EFL learners need to overcome. Being afraid of making mistakes, students tend to be reluctant to speak in the classroom. Teachers who constantly correct students' errors can intensify the students' apprehension. A key challenge in EFL learning is that how adequate help can be provided to learners with limitedteachers' intervention. This paper reports the design of a study on the use of text-to-speech (TTS) synthesis in English pronunciation learning 'with the aims to alleviate EFL students' language anxiety, and to empower them to learn pronunciation with limited teachers' support. The study adopts TTS synthesis as a pronunciation model in exercises focusing on the supra-segmental level. While coverage of the segmental level is not planned in the study, TTS synthesis facilitates learners to learn pronunciation of unfamiliar words by listening to the synthesised speech. The research question of this study is whether TTS synthesis can help alleviate EFL students' language anxiety, resulting in improving students' perception towards English pronunciation learning. Six one-hour sessions on English pronunciation are planned for a class of secondary 3 students over a two-month period. The instructional design is based on the John Keller's work which emphasises the motivation design. Data about the students' perception on English pronunciation learning will be collected in the first and the last sessions for analysis. Selected students will be asked for any perception change on English pronunciation learning in the study and the reasons behind in subsequent interviews.