Examining Factor-Level Changes in Teacher Feedback, Feedback Orientation, and Writing Across AI, Human, and Hybrid Feedback Conditions
Abstract (Research problem) This study addressed limited empirical examination of the multidimensional structures of teacher feedback and feedback orientation and of how their dimensions change under different AI exposure conditions in Chinese L1 writing classrooms. (Objectives) It tested the proposed five-factor teacher feedback and four-factor feedback orientation structures and compared pre–post changes across AI-only, AI–human, and human-only conditions. (Method) Parallel analysis, EFA, and CFA supported both structures; repeated-measures ANOVA tested condition differences. (Key findings) AI–human produced the largest gains on three feedback dimensions: verification, directive, scaffolding, all feedback orientation dimensions, and writing; AI-only produced moderate gains in the same dimensions; and human-only improved only praise and criticism. (Implications) Findings validate the dimensional structures, show that AI–human feedback constitutes higher-quality AI exposure, and inform differentiated integration of AI and human feedback to support specific feedback perceptions, feedback orientation, and writing improvement.
Authors
- Sze Yik Chu (ORCID: https://orcid.org/0009-0000-8193-3192)
- JianJun ZHU (ORCID: https://orcid.org/0009-0002-5114-2626)
Institutions
- Education University of Hong Kong (HK)
- Minnan University of Science and Technology (CN)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-16
- DOI
- https://doi.org/10.21203/rs.3.rs-11720044/v1
- Primary Topic
- EFL/ESL Teaching and Learning
- Type
- preprint