Automatic pronunciation evaluation feedback and peer correction for shaping English pronunciation accuracy and interpersonal communication
Maroua Rogti · Innovation in Language Learning and Teaching · 2025
EFL teachers in Algeria may inquire about the significance of offering specific strategies, such as corrective and personalized feedback for pronunciation training using Automatic Speech Recognition technology, which enables students to exert control over their language learning capabilities. This research aims to investigate the effect of combining an automated speech recognition system with peer correction to improve EFL students' pronunciation and enhance their interpersonal communication via peer correction. The research sample included 155 EFL students at the Higher College of Laghouat in the Fall of 2023. Participants were enrolled in a 10-week English language course, with 2 randomly allocated classes: the control group comprising 75 students and the experimental group including 80 students. CG students obtained ASR-based pronunciation evaluation feedback via the Speechace application, whilst EG students used the same tool with peer correction. The research used a quasi-experimental design including pre- and post-tests and one self-report questionnaire. The treatment included an evaluative component consisting of oral activities to test students' pronunciation. The analysis was conducted via a two-way repeated measures ANCOVA. The findings demonstrated that the use of ASR technology and peer correction significantly enhanced the pronunciation of Algerian EFL students and, therefore, improved their interpersonal communication skills. Educators and stakeholders make a preference for peer correction use on aspects of pronunciation training that would be affected by communication and cooperation or speaking elucidation. They subsequently could exploit the affordances of ASR technology. This can allow for obviating obstacles which can arise from explaining automatic feedback.