Speaker
Description
Fostering pragmatic speaking skills for high school seniors in Vietnam remains a stubborn pedagogical challenge, heavily constrained by large classes and grammar-centric, high-stakes exams. To address these bottlenecks, this study implemented a comprehensive "AI-assisted speaking ecosystem" tailored to the Global Success 12 curriculum during the 2026–2027 academic year. The framework integrates six target interventions: prompt-engineered material customization, 1-1 interactive simulations via AI Voice modes, pressure-free pronunciation self-correction through voice-to-text tools, differentiated task design, off-class virtual speaking journals, and generative multimedia projects.
A mixed-methods approach tracked 120 Grade 12 students to measure the framework's actual impact. Both quantitative and qualitative indicators revealed striking shifts in the classroom dynamic. Students’ speaking confidence and intrinsic initiative soared from a baseline of 16.7% to 79.2% post-intervention. Crucially, metrics tracking core oral proficiency—including fluency, accuracy, and spontaneous communicative reflex—all document improvements exceeding 50%. Rather than creating tech-dependency, this pedagogical scaffolding actively transformed passive learners into autonomous users who practice English with responsible digital literacy. The session presents a validated prompt library and a scalable instruction model that high school educators can immediately replicate in resource-limited EFL settings.
Keywords: Generative AI, EFL speaking, communicative confidence, Global Success 12, autonomous learning.
