ASR-Mediated Speaking Practice for Lower-Proficiency University EFL Learners: Perceptions across Novice and Beginner Levels

Although automatic speech recognition (ASR) has been increasingly used in L2 speaking instruction, relatively little is known about how lower-proficiency learners perceive sustained ASR-mediated speaking practice. This study investigated the perceptions of novice and beginner university EFL learners in general English courses. Survey data were collected from 252 undergraduate students enrolled in novice- and beginner-level classes at a South Korean national university. The analysis identified two factors: perceived value and overall evaluation of ASR-mediated speaking practice, and speaking preparation and self-directed practice. One item concerning perceived ease of use was analyzed separately. Responses were generally favorable for both factors and the separately analyzed item. No statistically significant differences were detected between the novice and beginner groups. Building on earlier classroom-based action research, the present study extends this line of inquiry to a larger cohort of lower-proficiency learners. Overall, the findings provide evidence of favorable learner perceptions in the present instructional context and support the feasibility of incorporating ASR-mediated speaking activities into university general English courses for lower-proficiency learners.