AI-assisted vocabulary instruction for IELTS candidates: A mixed-methods exploration

Abstract

This explanatory sequential mixed-methods study investigates the effectiveness of AI-assisted vocabulary instruction in improving the lexical range of International English Language Testing System (IELTS) candidates. Despite growing interest in Artificial Intelligence (AI) for educational purposes, research into its application within specific exam preparation contexts, such as IELTS, remains limited. Furthermore, there is insufficient exploration of AI’s impact on learners’ lexical range within short timeframes. To address these gaps, this study involved 40 IELTS candidates (aged 18–35, with band scores between 5.0 and 5.5) divided into experimental and control groups. Through a Vocabulary Size Test (VST) for pre-test and posttest, participants were assessed and quantitative data were gathered. The quantitative analysis revealed significant improvement in the experimental group. Descriptive statistics indicated a mean increase in vocabulary scores, with pre-test scores (M = 43.10, SD = 7.41) rising to posttest scores (M = 57.85, SD = 7.47). Qualitative data from semi-structured interviews identified three key themes: (1) perceived improvement in vocabulary, (2) increased engagement and motivation, and (3) challenges faced with AI tools. The findings demonstrate that AI-assisted vocabulary instruction can effectively enhance vocabulary development and motivate IELTS candidates, particularly those at low and intermediate proficiency levels. The study highlights both pedagogical implications for IELTS preparation and limitations related to sample size, instrument scope, and the role of teacher mediation.