A Quasi-experimental Study on the Effects of AI-assisted Peer Teaching on Vocabulary Retention among EFL Students in Vietnam

In response to the growing integration of artificial intelligence (AI) in education, language instructors face challenges in effectively blending AI tools with pedagogical practices, particularly for enhancing vocabulary acquisition among non-English major learners. This study investigates the impact of AI-assisted peer teaching on vocabulary retention and explores 42 students’ perceptions of AI-generated materials within a peer-led learning environment. Employing a quasi-experimental repeated-measures design, the study tracked vocabulary development across four formative post-tests among second-year students at Ho Chi Minh University of Banking (HUB). Results revealed a significant upward trend in retention, with mean scores increasing from 6.71 (SD = 0.87) to 8.84 (SD = 0.72), supported by repeated-measures ANOVA (F = 186.34, p < 0.001, η² = 0.820). Survey responses indicated strong student approval of AI-supported learning, with 76.2% expressing positive perceptions towards this framework. The study also revealed that future research should consider incorporating longitudinal designs, refining AI content for specialized contexts, and broadening the sample size to enhance generalizability in AI-assisted vocabulary instructions.