ASR using Speechnotes for EFL Learners: A Study of the Effects on English Pronunciation and Prosody Skills

Abstract

Artificial Intelligence has captured the language classrooms all over the world like never before, albeit not without many debates and discussions on the pros and cons. However, there exists some kind of consensus on the many opportunities that AI powered applications offer to the learners. This study concerns itself with the role of Automatic Speech Recognition technology alongside peer feedback on the English language (EFL) pronunciation and speaking skills of adult Saudi learners. The study uses a mixed-methods approach and an exploratory sequential design to examine whether and what is the impact of an AI application on the EFL output and whether such technology can be integrated as a learning aid in the Saudi higher education realm. The participants are 24 intermediate proficiency Saudi EFL learners at University of Hai enrolled in the sophomore year of the English Speaking Skills Course. Prior to the study, they had only received conventional instruction in pronunciation before being subject to the Speechnotes ASR application. The impact of the intervention on pronunciation and general speaking proficiency of prosodic components was studied and judged in the specially designed tasks and the standardized IELTS English Speaking Skills component. Qualitative data in the form of semi structured interviews with the sample helped triangulate the information to answer the main point of inquiry in this study: Can an ASR tech tool be effectively used as a supportive pronunciation and general speaking skills learning aid in an EFL class of university learners.