.png){kind=icon}
.png){kind=icon}
Return
University student agency in genre-sensitive writing: a longitudinal mixed-methods multi-case study of teacher and GenAI feedback
.png){kind=icon}
OA
AI
H
L
D
DOI：10.1080/03075079.2026.2691927.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Learner agency is central to effective writing, and yet little is known about how students mobilize evaluative judgement when coordinating teacher feedback with feedback from generative artificial intelligence (GenAI) tools. This longitudinal mixed-methods multiple-case study (N = 10) traced how EFL university learners engaged with hybrid instructor–GenAI feedback across three genre-sequenced tasks (argumentative essay, AE; analytical report, AR; research proposal, RP) and three successive revision rounds (D1→D2→D3) within each task. We identified T = 403 actionable feedback points and defined item-level uptake (U) as the number of actionable points realized in later drafts. Overall, U = 224 were adopted (U/T = 55.6%). Source-specific analyses showed higher uptake for teacher-only feedback than for GenAI-only feedback (61.9% vs. 45.5%), while convergent teacher-GenAI feedback showed the strongest uptake tendency (61/74 = 82.4%). Revision-depth analyses showed task-sequenced variation rather than a simple linear developmental shift: AR showed the strongest concentration of surface-level edits (71.0%) and the highest revision density, whereas AE and RP involved relatively higher proportions of meaning-level edits. Sensitivity analyses using alternative handling of convergent and ambiguous cases did not alter the main source-category pattern. Qualitative joint displays of prompt-response logs, reflective journals, and interviews indicated that learners calibrated their use of teacher and GenAI feedback by evaluating, adapting, rejecting, or sequencing suggestions according to epistemic risk and task demands. The findings show that teacher and GenAI feedback functioned as complementary epistemic resources and that learner agency was reflected in source-sensitive, genre-sequenced evaluative coordination rather than in feedback uptake alone.
Keywords:
Feedback uptake
genre-sensitive academic writing
human-AI collaboration
learner agency
evaluative judgement
Journal
IF:
3.2
Papers:
3.3K
Citations:
1.0W
Organization
Cited Papers