Hanxu Tu, Wan Zhou

Hanxu Tu, Wan Zhou

Oct. 30, 2026

AI-supported speaking, Vocational EFL,

China’s English Curriculum Standards for Higher Vocational Education (2021) lists workplace-oriented communication among the core competencies of vocational English education. Large classes and limited contact time, however, make sustained individual speaking practice difficult. This study examined how students’ practice behaviors were associated with speaking gains and whether gains differed between pronunciation and communicative effectiveness. It also tested whether baseline speaking proficiency changed the association between total attempts and gain. A semester-long single-group pretest–posttest classroom study involved 99 second-year students in a 16-week General Workplace English course. Students used an AI-supported speaking practice platform, while parallel teacher-rated tasks measured overall speaking, pronunciation, and communicative effectiveness. Platform logs provided total attempts and repeat frequency. Overall speaking scores increased from 62.30 to 70.40, a mean gain of 8.10 points. Both practice indicators were positively associated with overall speaking gain. The association between attempts and gain strengthened as baseline proficiency increased. The simple slope was weak and nonsignificant at one standard deviation below the baseline mean, but positive at the mean and one standard deviation above it. Pronunciation and communicative effectiveness both improved, with pronunciation gaining 1.92 points more on matched 20-point scales. Teachers can read practice logs alongside baseline proficiency to identify participation patterns and plan support. In vocational English courses, platform work can provide frequent rehearsal and pronunciation practice, while classroom interaction can develop extended responses, role-play, and meaning negotiation.

© 2026, the Authors. Published by ISCCAC

This is an open access article distributed under the CC BY-NC license