Most student essays remained at a descriptive reflection level despite repeated AI feedback¶

Claim · All claims
Evidence · 1 study · associational r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? 94% of essays were coded at the descriptive reflection level; dialogic reflection occurred in 5% of essays and critical reflection in less than 1%. → Andres Karjus 2026

Evidence¶

Andres Karjus 2026¶

Andres Karjus, Janika Leoste, Tiia Õun. (2026). Student Evaluation of Repeated AI Feedback Across a Semester of Writing. arXiv:2607.16115 [cs.CY]. https://arxiv.org/abs/2607.16115 doi:10.48550/arxiv.2607.16115

q2 · i? · associational · r2

Coding of essay reflection depth with a Hatton–Smith-derived four-category scheme in the semester corpus. Most coded essays were descriptive reflection, connecting course content to personal experience but rarely moving into sustained alternative perspectives or critique.

"Dialogic reflection occurred in 5% of essays, and the deepest category, critical reflection, in less than 1%."

Discussion¶

Related Claims¶

• Expressed helpfulness of AI feedback showed no association with next essay reflection depth — related
• Students found AI feedback helpful and actionable more often than not across a semester of reflective writing — related
• Student appraisals became more critical over the semester, with actionable-support themes falling from 67% to 43% — related
• In the graduate course, most individual synthesis artifacts remained at surface level, with format and depth varying by group and reflecting differing conceptions of synthesis — related
• Agency erosion appeared in 88% of student essays, most often as erosion of independent effort — related
• GenAI's default output on a take-home exam question remained at the quantitative, multistructural SOLO phase, producing polished but pedagogically thin responses — related