AI-ASSISTED POST-SPEAKING REFLECTIVE STRATEGIES: EVALUATING METACOGNITIVE EVALUATION TOOLS FOR EFL FLUENCY DEVELOPMENT

Abstract

Developing oral fluency in English as a Foreign Language (EFL) requires not only continuous practice but also structured metacognitive evaluation of one's speech. While interactive voice AI tools offer accessible environments for real-time oral practice, relying solely on conversational interaction often fails to address recurring speech hesitation and structural errors. This paper investigates the pedagogical impact of integrating AI-assisted post-speaking reflective strategies among undergraduate EFL learners. Utilizing a mixed-methods action research framework, the study examines how AI-generated speech transcripts, audio recordings, and structured self-evaluation worksheets support fluency development. The findings demonstrate that post-task metacognitive reflection significantly enhances students' awareness of speech disfluencies, leading to a noticeable reduction in hesitation markers and unnatural pauses during subsequent speaking tasks. Furthermore, the visual evidence provided by AI transcripts empowers learners to make targeted self-corrections and expand their vocabulary independently. The study concludes that combining interactive voice AI platforms with structured post-speaking metacognitive evaluation creates an effective framework for fostering oral fluency and autonomous learning in modern EFL classrooms.