Abstract:

Despite growing interest in AI-assisted language learning, empirical evidence on AI-generated instructional videos for context-specific vocabulary acquisition remains insufficient. This study examined the effectiveness of AI-generated instructional videos in supporting vocabulary acquisition among English as a Foreign Language (EFL) learners in a workplace English course at the university graduate level. An explanatory sequential mixed-methods design was employed with 161 EFL learners, combining a pretest, a delayed posttest, and a learner-perception survey.

Quantitative results showed a statistically significant improvement in vocabulary knowledge, with a mean pretest score of 7.1 (SD = 1.33) increasing to 9.2 (SD = 0.89) at posttest, and a large effect size, Cohen's d = 1.62. Qualitative findings revealed predominantly positive learner perceptions, particularly regarding clarity, multimodal comprehensibility, and the instructional value of workplace-contextualized vocabulary presentation. Participants specifically highlighted how visual and auditory elements supported their understanding of professional vocabulary in meaningful situations.

These findings suggest that AI-generated instructional videos designed around authentic professional scenarios may support vocabulary development in workplace-oriented EFL contexts. The study contributes empirical evidence to the emerging field of AI-generated learning materials and offers practical implications for technology-integrated vocabulary instruction in workplace English courses.

Keywords:

AI-generated videos, EFL, vocabulary acquisition, learner perceptions, workplace English, educational technology.