When Scaffolding Breaks: Investigating Student Interaction with LLM-Based Writing Support in Real-Time K-12 EFL Classrooms
Junho Myung, Hyunseung Lim, Hana Oh, Hyoungwook Jin, Nayeon Kang, So-Yeon Ahn, Hwajung Hong, Alice Oh, Juho Kim
CHI 2026 · Accepted
Deploying LLM writing support in real K-12 classrooms revealed that the most revealing moments were when the AI failed — exposing how students had been relying on it all along.
My contributions
- Longitudinal classroom deployment study of LLM writing support across real K-12 EFL lessons
- Taxonomy of student responses to AI scaffold failure
- Analysis of how reliance patterns differ from what surface-level engagement metrics suggest