Japanese EFL Learners' Uses of Text-to-Speech Technology and Their Learning Behaviors: A Pilot Study

Abstract

The purpose of this study is to show how EFL learners work on Text-to-speech (TTS) technology to improve their speaking accuracy. We constructed the system for improving speaking proficiency based on Web-Speech API. Usually, text-to-speech is available on the net or by using TTS applications in education. However, these sounds produced are not generally customized to the diversity of learners' learning strategies. The system constructed in the study makes it possible for learners to choose the words, phrases sentences or the passage they like to listen to for their speaking practice. Major findings that came out from this study are: (i) As to the easier materials, accuracy improvement was observed regardless of their proficiency, (ii) On the contrary, as to the difficult materials, such an improvement was not observed in the experiment, and (iii) there is diversity of how they worked on the system, and there are some patterns in their learning behaviors, Bottom-uppers and Top-downers. These findings suggest the usefulness of the current system for various types of learners in terms of learning strategies and a future possibility from a learning analytic point of view to modify the system.