<![CDATA[Pronunciation plays a pivotal role in oral communicative competence; however, it remains underemphasized in many Indonesian EFL classrooms due to time constraints and limited individualized feedback. This study investigates the effectiveness of SpeechAce, anArtificial Intelligence (AI)-based Automatic Speech Recognition (ASR) platform, in improving pronunciation accuracy among eighthgrade students at SMPN 4 Sambas. A pre-experimental one-group pretest and post-test design was employed involving 28 students. Quantitative data were collected through AI-generated pronunciation scores. Statistical analysis revealed a substantial improvement from a mean pre-test score of 53.50 to a mean post-test score of 83.36. The computed t-value (16.77) exceeded the critical t-table value (2.052) at α = 0.05. The effect size (Cohen’s d = 3.15) indicates an exceptionally strong practical impact. Frequency distribution analysis further demonstrates a complete elimination of low-score categories and significant upward mobility into very high and excellent performance bands. The findings confirm that AI-based pronunciation assessment significantly enhances pronunciation accuracy and supports masteryoriented learning.]]>
Copyrights © 2026
