Exploring the Use of AI-Generated Contextual Sentences for English Vocabulary Acquisition Among Malaysian ESL Learners

Abstract

This study investigates the efficacy of AI-generated contextual sentences in enhancing vocabulary acquisition among Malaysian ESL learners. Vocabulary mastery is pivotal for language proficiency, yet rote learning methods often fail to provide the necessary context for long-term retention. Adopting a mixed-method research design, this study utilises methodological triangulation comprising pre- and post-tests, semi-structured teacher interviews, and student interviews. The study involved 10 Form Four ESL students and 3 English teachers from a Malaysian secondary school in Ipoh, Perak, and evaluated the effectiveness of AI-customised vocabulary exercises tailored to learner interests. Quantitative data were analysed using pre- and post-test comparisons, which revealed a clear improvement in vocabulary definition accuracy and contextual usage following the intervention. Qualitative findings further indicated heightened learner engagement and motivation. However, discrepancies between written vocabulary mastery and oral fluency were observed. These findings provide important implications for the pedagogical integration of Generative AI in English Language Teaching (ELT).