Artificial Intelligence in Flipped Classrooms for Enhancing Vocabulary Acquisition and Retention in Primary ESL Education

Abstract

Persistent difficulties in vocabulary mastery, particularly in balancing receptive and productive knowledge with long-term retention, remain a challenge among Malaysian primary learners despite ongoing curriculum reforms. This study investigated the effectiveness of the Vocabulary Flipped Classroom Model (VFCM) in improving vocabulary outcomes among Year Three Malaysian learners of English as a Second Language (ESL). The study aimed to examine whether an artificial intelligence (AI)-enhanced flipped classroom approach could significantly improve learners’ receptive vocabulary knowledge, productive vocabulary knowledge, and vocabulary retention compared with traditional teacher-centred instruction. The study employed a quasi-experimental pre-test–post-test design with a delayed post-test, involving 80 learners who were divided equally into experimental and control groups. The intervention integrated structured pre-class vocabulary tasks supported by ChatGPT, guided in-class practice, and systematic corrective feedback. Results indicated that the experimental group demonstrated significantly greater gains in both receptive and productive vocabulary, with moderate to large effect sizes, and sustained stronger retention over time than the control group. These findings highlight the pedagogical value of embedding AI-supported flipped learning within primary English classrooms to move beyond rote memorisation toward deeper lexical understanding and durable learning. The study contributes empirical evidence to current discussions on technology-enhanced language instruction and offers a scalable instructional framework for teachers, curriculum planners, and policymakers seeking sustainable vocabulary development strategies.