Read it twice, a day apart
Peer review comments are compressed, often written quickly, and rarely softened. A remark such as “the analysis is inadequate” usually means one specific thing the reviewer did not have space to elaborate, but on first reading it lands as a verdict on the work. The gap between those two readings is why the twenty-four hour rule is worth keeping.
On the second pass, translate each comment into the underlying request. “The introduction is unfocused” often means the research question is not stated explicitly. “More detail on the sample is needed” usually means a specific number or criterion is missing. Working out what would satisfy the comment is more productive than assessing whether it is fair.
It also helps to notice what a review implies about the paper’s standing. A long, detailed review is usually a sign that the reviewer thinks the work is worth improving. Editors rarely invite a revision for papers they intend to reject, so a major-revision decision is closer to good news than it feels on the day it arrives.
The anatomy of a response letter
The letter has a short frame and a long body. The frame is two or three sentences: thank the editor and reviewers, state that you have revised the manuscript, and note that responses follow point by point with locations. One thank-you is sufficient; repeating it before every response reads as filler and makes the letter harder to scan.
The body reproduces every comment in order, grouped by reviewer, each followed by your response. Use consistent typographic distinction — reviewer comments in one style, your responses in another — so the editor can move through it quickly. Number your responses to match the reviewer’s numbering, and where a reviewer wrote a wall of prose, split it and say that you have done so.
Where two reviewers raise the same point, answer it fully in the first instance and cross-reference from the second rather than pasting the same paragraph twice. Editors read the whole letter and notice repetition; reviewers read their own section and need it to be complete. When you cross-reference, give the reviewer number and the comment number so the pointer can be followed.
What a good response to a single comment contains
Three elements: acknowledgment, action, and location. “We agree that the sample description was too brief. We have added inclusion and exclusion criteria and the recruitment window (Methods, p. 7, lines 142–158).” That is complete, checkable, and takes one sentence more than the useless version.
The useless version is “this has been addressed”, with no indication of what changed or where. It forces the reviewer to search the manuscript, and reviewers under time pressure who cannot find a change sometimes conclude it was not made. Every response should be verifiable without hunting.
For substantial changes, quoting the new text in the letter saves everyone time. Keep the quotation to the sentences that answer the comment rather than reproducing whole sections. Where a change ripples through the paper — a reanalysis that alters several numbers — say so explicitly and list the sections affected.
Disagreeing without losing the paper
You are not obliged to make every requested change, and editors do not expect you to. What they expect is a reason. The pattern that works is: acknowledge the concern, explain why you have taken a different course, and offer whatever partial accommodation is possible — a clarification in the text, a limitation added to the discussion, a supplementary analysis in an appendix.
Keep the reasoning about the evidence rather than about the reviewer. “We appreciate the suggestion. A mediation analysis would require repeated measures, which this cross-sectional design does not provide; we have added a sentence in the limitations noting that mediation remains untested (p. 19).” That is a decline, and it is very unlikely to cause offense. Compare “the reviewer appears to misunderstand the design”, which says the same thing and invites a fight.
Where a reviewer is factually wrong, correct the record gently and consider why the misreading happened. If a competent reader misunderstood your design, the manuscript probably invited it, and a clarifying sentence serves you better than winning the point. Editors notice authors who improve the paper in response to a mistaken comment.
Comments you cannot address
Some requests are impossible within the study: another year of follow-up, a population you cannot access, an experiment that would be a separate paper. Say so directly, explain the constraint, and state what you have done instead — usually acknowledging it as a limitation and specifying it as future work. Editors accept this routinely when the reasoning is clear.
Requests that would change the paper into a different one deserve the same treatment. Reviewers sometimes describe the study they would have run. Where the suggestion is genuinely a different project, say that you agree it is a valuable question, that it falls outside the scope of the present analysis, and where you have noted it in the discussion.
When reviewers contradict each other, do not try to satisfy both. Choose the course you can defend, say explicitly that the two reviewers took different views, explain your reasoning, and let the editor adjudicate. That is precisely the decision the editor is there to make, and flagging it is helpful rather than evasive.
