AI-assisted feedback drafts increase discretionary feedback provision by TAs by 10.81 percentage points¶

Claim · All claims
Evidence · 1 study · causal r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? Treatment assignment increased feedback provision by 10.81 percentage points (SE = 1.10, p < 0.001) in a randomized field experiment where feedback was optional. → Mahinpei 2026

Evidence¶

Mahinpei 2026¶

Mahinpei, R., Dean, V., Fong, R., Liu, L. T., & Ribeiro, M. H. (2026). AI Assistance for Discretionary Work: Increasing Feedback Provision in Higher Education. arXiv. https://arxiv.org/abs/2606.03095 doi:10.48550/arxiv.2606.03095

q3 · i? · causal · r2

Pre-registered semester-long randomized field experiment in a 300-level ML course with 11 TAs and 88 students; question-level randomization with rotating assignment across four homework assignments. Within-question linear probability model with homework-question fixed effects estimated a "10.81 percentage points" increase in feedback provision (p < 0.001).

"We find that treatment assignment increased feedback provision by 10.81 percentage points (SE = 1.10, 95% CI [8.60, 13.00], 𝑝< 0.001)"

Discussion¶

Related Claims¶

• AI-assisted feedback does not reduce student-perceived usefulness of rated feedback — related
• AI-assisted feedback drafts do not reduce TAs' time spent per character of feedback — related
• AI-assisted feedback drafts increase total feedback length by 39.79 characters per submission — related
• TAs adapt rather than copy AI feedback drafts, with moderate draft–final overlap (ROUGE-L F1 = 0.60) — related
• Draft usefulness varies by grading context: higher for submissions with mistakes and applied questions — related
• In the Enacted Feedback workflow most resources moved from AI feedback to suggestion selection, while optional AI assistance was rarely entered — related