How do I respond to peer reviewer comments?

Wait before you write anything. The first reading of a review is emotional and the letter you would write in that hour is not the letter you want to send. Two or three days later the same comments read as a list of tasks, and most of them turn out to be reasonable.

Write two documents, not one. Noble recommends an initial private document where you and your coauthors interpret the comments, compare possible responses, weigh the cost and benefit of extra experiments, and vent frustration, then a separate professional response written later, after some time has passed and after you have started addressing the critiques (Noble, 2017, pp. 3-4). The private draft is where the annoyance goes. The submitted letter is where the reasoning goes.

Then sort the comments. Some are straightforward fixes. Some ask for analysis you can do. Some ask for something impossible, such as data you cannot collect. Some rest on a misreading. Each category needs a different response, and doing this sorting before writing keeps the letter consistent in tone.

The letter's job is to let the editor decide quickly. That means quoting each comment, responding directly underneath it, and pointing to the exact place in the manuscript where the change appears. An editor who has to cross-reference three documents to check one point will take longer and be less inclined to accept.

Open with a summary of the major changes before you reach the point-by-point section, leading with new data or analyses that address the most important criticisms, and note where several reviewers raised the same issue. Then reproduce the complete reviews with your replies interleaved (Noble, 2017, p. 1). Inside that section, a three-part unit works: the reviewer comment, your response, and the change in the manuscript, with typography that keeps the three visually distinct (Noble, 2017, p. 2). Answer every point, minor comments included, rather than only the ones that are easy to answer, and make each reply self-contained so the reviewer does not have to hunt through the revised manuscript to see what you did (Noble, 2017, p. 2).

Thank the reviewers once, briefly, at the start. Extended gratitude in every response reads as anxiety. What actually persuades is evidence that you engaged with the substance.

What do I do with a reviewer who is simply wrong?

Disagree politely, give the reason, and treat the misunderstanding as evidence that your text was unclear. Then change something. A response that pushes back on every point without a single revision reads as defensive even when you are right on the substance.

The framing that works is that the reviewer read carefully and reached a conclusion your text permitted. Even when the reading was careless, writing it this way keeps the exchange productive and costs you nothing.

Noble's version of this is to accept that the presentation failed to communicate the point, revise the passage, and quote the revised wording directly in your response so the reviewer can see the fix without opening the manuscript (Noble, 2017, p. 2). That is not the same as conceding the point. It is worth knowing that reviewer judgements are less stable than the confident tone of a report suggests: Tennant and Ross-Hellauer report that studies of inter-reviewer reliability give variable results and that the research base on peer review is theoretically and empirically underdeveloped, with conflicting or inconclusive findings (Tennant and Ross-Hellauer, 2020, p. 1). They also note that reviewers usually cannot reproduce the submitted experiments and so have to trust your reported procedures, data, and analyses (Tennant and Ross-Hellauer, 2020, p. 7), and that impartiality can be affected by author nationality, institutional prestige, gender, discipline, confirmation bias, and workload, though much of that evidence comes from small journal-level datasets (Tennant and Ross-Hellauer, 2020, p. 4). Respond to every point, but separate a required correction from a preference.

Where a reviewer asks for something genuinely outside the scope of the paper, say so and explain the boundary. Editors are used to arbitrating this and will usually accept a clear, reasoned limit.

How do I structure the response letter?

A short covering paragraph, then each comment quoted in full, numbered, with your response immediately beneath it and the revised text quoted or the page number given. Reviewer by reviewer, in their order. Nothing clever, and nothing an editor has to search for.

Use formatting to separate comment from response, since a wall of undifferentiated text is hard to read. Bold or italic for the comment, plain for your reply, and the revised manuscript text indented.

Submit a tracked-changes version alongside a clean one unless the journal says otherwise. Editors and returning reviewers use it to check the revisions in minutes, and providing it makes acceptance faster.

When a reviewer's request contradicts an editor's instruction, hand the conflict to the editor instead of resolving it yourself. Noble's example is a reviewer asking you to move supplementary material into the main text while the editor has asked you to shorten the article, and the advice is to state the conflict and leave it to the editor's discretion (Noble, 2017, pp. 3-4). The same section asks you to identify and explain your changes relative to the previous version, which is what lets an editor verify the revision instead of taking your word for it (Noble, 2017, pp. 3-4).

Check the deadline and ask for an extension early if you need one. Most editors grant a few weeks without comment, and a late revision that arrives unannounced is treated as a new submission at some journals.