We are conducting a systematic review and meta-analysis on AI feedback and its effects on human performance. The goal is to synthesize evidence on whether and when AI-generated feedback improves human outcomes, and to identify conditions under which it may fail to help or even backfire. To reduce publication bias and ensure a comprehensive review, we are looking for unpublished studies, in-press papers, or work under review that fit the following criteria:
- The study includes human participants completing a task for which they receive AI-generated feedback
- The feedback may be presented in any format, including textual, visual, numerical, or multimodal forms.
- The study reports at least one quantitative performance outcome, such as accuracy, error rate, quality ratings, task completion speed, or creativity.
If you have a relevant unpublished or in-press manuscript, or a study under review, we would be very grateful if you could share it with us. You can submit the manuscript/data via email: y.luan@uq.edu.au
We kindly ask for submissions by August 23, 2026.
We would also greatly appreciate it if you could share this call with colleagues or others conducting related research.
Thank you for helping us build a more complete and inclusive evidence base on AI-generated feedback and human performance.
Yeun Joon Kim, PhD
*On behalf of the author team: Dr. Luna Luan, Dr. Jungmin Choi, Mr. Pengzhao Lyu, Professor Soo Min Toh, and Ms. Nayoung Kim
------------------------------
Yeun Joon Kim, PhD.
Associate Professor
Judge Business School.
Institute of Metabolic Science, School of Clinical Medicine.
University of Cambridge
Email:
y.kim@jbs.cam.ac.uk------------------------------