Speaker
Description
This study examines the effectiveness of the PEEL/PETAL framework in improving paragraph writing among Grade 10 students in a Vietnamese EFL context under the Global Success curriculum. It addresses persistent challenges such as fragmented ideas, weak coherence, and limited use of supporting evidence. The study employed a mixed-methods action research design conducted during the 2025–2026 academic year. The intervention involved a five-step pedagogical cycle, including model analysis, scaffolded planning, and peer feedback across multiple writing genres.
Quantitative findings indicate substantial improvements in students’ logical organization (+44%) and evidence integration (+42%), alongside a notable increase in writing self-efficacy. Qualitatively, the framework successfully reduced L1 interference and cognitive load by separating idea organization from linguistic production.
Importantly, in the context of increasing AI-assisted writing tools, this study highlights the role of pedagogical scaffolds in strengthening learner agency and reducing over-reliance on generative technologies. By providing a clear cognitive structure, PEEL/PETAL enables students to critically engage with content rather than passively depending on AI-generated outputs.
The study contributes to current discussions on AI literacy and assessment by demonstrating how traditional scaffolding strategies can be adapted to support responsible and effective writing practices in AI-enhanced EFL classrooms.
Keywords: PEEL/PETAL, EFL writing, AI-assisted learning, learner agency, evidence-based writing, action research
