Two author teams revise their papers with equal care. The first sends a ten-page response: every comment quoted, answered clearly, with page and line numbers and the new text. The second writes “We have revised the manuscript accordingly” after each comment. The first paper is accepted after round two. The second receives another major revision, because the reviewer could not find the changes and concluded the authors had not made them.
The response to reviewers is the document reviewers read first in round two, often before the manuscript itself. Its job is to make checking your revision so easy that a reviewer can tick “addressed” without having to search.
Overall structure
- A short opening to the editor: one specific sentence of thanks and a summary of the three to five biggest changes.
- Responses to each reviewer, in their order and numbering.
- Other changes the authors made on their own initiative, such as a corrected figure or updated references, stated plainly so nobody is surprised.
For each comment use three parts: the comment (quoted verbatim, in italics or indented), the response (a concise explanation) and the change (page, line and the new wording). Many authors put responses in a different colour; make sure the layout still works when printed in black and white.
Show exactly where things changed
“Revised” is not enough. The reviewer needs to know where and how. Write: “We have added a sensitivity analysis excluding the 34 respondents with missing data (p. 14, lines 312–320). Results were essentially unchanged (Table S4).” and quote the new passage underneath.
A tip from people who do this often: page and line numbers shift constantly during revision, so fill them in only at the very end, once the manuscript is final. While revising, drop a tag such as [R1.3] into the manuscript to find each change again.
Journals usually also want a marked-up version, with tracked changes or highlighting. If you have rewritten most of a section, tracked changes become a wall of red; highlight the new text instead and say in the letter that the section was rewritten in full.
Phrasing for common situations
| Situation | Suggested phrasing |
|---|---|
| Agree and revised | We agree and have revised the paragraph to clarify… (p. 8, lines 150–158). |
| Partly agree | We agree that… However, because…, we have instead… |
| Reviewer misread | We apologise that this was unclear. We have rewritten… to make explicit that… |
| Disagree | We respectfully disagree, for the following reasons… We have added a sentence acknowledging this alternative view (p. 12, lines 240–243). |
| Cannot be done | Unfortunately, these data were not collected in our study. We now acknowledge this as a limitation (p. 19, lines 401–405). |
| Out of scope | This is an excellent suggestion for future work, which we now mention in the Discussion (p. 20, lines 420–423). |
Note the “we apologise that this was unclear” move: it places the problem in the text rather than in the reader, and it is nearly always true. If an expert misread it, other readers will too.
Disagreeing politely without giving ground
Disagreeing with a reviewer is normal, but it needs reasons and evidence. A good rebuttal usually has four steps:
- Acknowledge what is reasonable in the comment.
- Give the scientific reason you are not following it: design, assumptions, literature.
- Provide evidence: citations, a supplementary analysis, a diagnostic check.
- Make a small change that shows you listened, often a sentence in the discussion or limitations.
Example: a reviewer suggests a multilevel model. You have 18 clusters, too few to estimate cluster-level variance reliably. Explain this, cite the methodological literature, show results with cluster-robust standard errors and add a sentence to the limitations. Reviewers rarely push further when they see the issue handled seriously.
Mistakes that drag out the second round
- Explaining in the letter but not changing the manuscript. The reviewer asked because the text was unclear; a private explanation does nothing for future readers.
- Answering several comments with one combined reply, leaving the reviewer to match them up.
- Skipping small points such as typos or figure legends. Reviewers often tick off every item and notice immediately.
- Thanking the reviewer in every single response. One specific thank-you at the top is enough; repeating “we thank the reviewer for this insightful comment” forty times reads as mechanical.
- Sarcasm, or implying the reviewer lacks expertise. The editor reads the whole letter and will note the tone.
In practice: three read-throughs before sending
- Cross-check: print your revision plan and tick off each comment ID that has a response in the letter.
- Location check: open the final manuscript and follow every page and line reference in the letter to confirm the text is really there.
- Hostile read: ask a co-author who did not do the revision to read the letter and mark anything that sounds defensive or vague.
The three passes take about half a day, which is trivial next to another round of review lasting months.
Next step: create the response document the moment you build your revision plan, paste in every comment verbatim with its ID, and fill in replies as you revise the manuscript. Leave page and line numbers for the final day.
Câu hỏi thường gặp
How long should a response to reviewers letter be?
There is no fixed length; it depends on the number of comments. What matters is that each comment is quoted, answered and linked to the exact change, even if the letter runs to ten pages.
Should I copy the reviewer comments into my response?
Yes, quote each comment verbatim before your reply. It lets reviewers and editors check your answers without switching back to the original report.
How do I disagree with a reviewer politely?
Acknowledge what is reasonable, give your scientific reasoning, provide evidence and make a small change such as a sentence in the discussion. Keep the tone neutral and avoid sarcasm.
Should I submit tracked changes or a highlighted manuscript?
Follow the journal’s instructions. If you rewrote most of a section, highlighting new text is usually easier to read than dense tracked changes.
Do I need to thank the reviewer in every response?
No. One specific thank-you at the start of the letter is enough. Repeating it in every reply lengthens the letter and sounds mechanical.
