VietTESOL International Convention 2026

Implementing Vietnam’s 2018 General Education Program requires a strong focus on communicative competence, yet massive class sizes make individual speaking assessments nearly impossible. This project addresses two research questions: how does an AI-assisted flipped classroom model affect secondary students' oral proficiency, and to what extent does it optimize teacher workload? Drawing on computer-assisted language learning (CALL), the pedagogy follows a four-step routine where students read Global Success texts at home via Microsoft’s Reading Coach to receive instant AI feedback, allowing the teacher to use Insights dashboards for targeted in-class remedies. Employing a mixed-methods longitudinal design over two years across grades 6–9 and 10–11 with students from three different secondary schools, the research tracked student test scores, system data, and grading logs. Findings revealed clear progress, including a 90% star-reward completion rate for grade 6, a 20% jump in vocabulary scores for older groups, and a significant reduction in foreign language speaking anxiety. However, the data also indicated that teachers’ and students’ AI literacy is crucial, as scoring reliability requires careful verification and further investigation regarding regional accents and background noise. Crucially, the model optimized teacher workload by cutting oral grading time by 40%. Attendees will receive actionable, effective strategies and ready-to-use lesson templates to replicate this digital ecosystem in their own under-resourced classrooms.