A reviewer report is not a verdict, it is a to-do list. Sort every comment into required and recommended, answer each one in a numbered response letter that says what you changed and where, and make the changes in the manuscript with tracked changes. You may disagree — but you have to answer with a reason, not silence. Unanswered comments are the single commonest reason a revision comes back again.
The email arrives and it is three pages of criticism. This is normal, it is not a rejection, and the way you handle the next fortnight largely decides whether the paper gets published.
A reviewer report is a to-do list with a deadline, and it should be processed like one.
First, read it twice and do nothing
Read it once for content. Then close it for a day.
First reactions to peer review are reliably bad — either defensive, or disproportionately discouraged. Neither state produces a good revision. The report will read differently on the second pass, and the comments that felt like attacks usually turn out to be requests for information you already have.
What you are looking for on the second read is structure: how many distinct asks are there, and which ones change the paper versus change the prose?
Sort every comment into three piles
| Pile | What it is | What you do |
|---|---|---|
| Required | The journal says it must change | Make the change |
| Recommended | The reviewer suggests it | Make it, or decline with a reason |
| Misreading | The reviewer has misunderstood something | Clarify the manuscript, then explain |
Good journals do this sorting for you. The Journal of Emerging Investigators returns an editor's letter setting out reviewers' requested changes labelled as required or recommended and organised by manuscript section — which is exactly the structure you want. Where a journal does not label them, sort them yourself and say in your letter how you have read each one.
The third pile deserves attention, because students handle it worst. If a reviewer has misunderstood your method, that is usually evidence the method section is unclear — the reviewer is an expert reader who still got it wrong. The fix is to improve the manuscript and then explain, not just to explain.
Write a numbered response letter
This is the deliverable, and most students do not know it exists.
The format, which is standard across academic publishing:
Reviewer 1, comment 3. "The sample is not described in enough detail to judge generalisability."
We agree. Section 3.1 now states the recruitment method, the year groups sampled, the response rate (62%) and the two exclusions, and the limitations section notes that the sample is drawn from a single school and should not be generalised beyond comparable settings. See page 7, paragraph 2.
Three things that entry does:
- Quotes the comment, so the editor does not have to cross-reference.
- Says what changed, specifically.
- Says where, by section and page.
Do that for every comment, in order, including the ones you have declined. An editor checking a revision is looking for one thing above all: has every point been addressed? Completeness beats elegance.
How to disagree properly
You are allowed to push back. Handled well it strengthens your standing with the editor; handled badly it is why papers get rejected on the second round.
The structure that works:
Reviewer 2, comment 5. "The authors should run a regression controlling for prior attainment."
We considered this. Prior attainment data were not available for this cohort — the school does not release them for research use, which Section 3.2 now states explicitly. We have instead reported results separately by year group (Table 4), which addresses the underlying concern about composition without requiring the unavailable variable. We have added this constraint to the limitations.
Note what it does. It acknowledges the point, gives a reason that is external to your preferences, offers a substitute that addresses the concern, and concedes the residual weakness.
What does not work: silence, restating the original claim more confidently, or "we feel this is outside the scope of the paper" with nothing after it. Scope is a legitimate reason; it just has to be argued.
Make the changes visible
Revise in the manuscript with tracked changes, unless told otherwise. JEI's copy-editing stage works exactly this way — editors make line edits and comments in the manuscript file, and authors review and revise with track changes.
Submit both: the clean revision and the response letter. If the journal wants a marked-up version too, send all three. Never submit a revision without the letter; it reads as though you have ignored the review.
Know when to stop
There is a point beyond which you stop improving the paper, and journals mark it. JEI's proofing stage is explicitly for strictly necessary corrections, not further copy editing — a proofing editor produces the final PDF and you check it for errors, not for things you would now phrase differently.
Respect that line. Authors who keep rewriting at proof stage delay their own publication and irritate the people who have just spent months on their paper.
Timing
Your decision email states a deadline. Meet it, or ask for an extension before it passes.
Two things worth knowing about journal time. Review is multi-stage and slow — a manuscript may pass back and forth with a managing editor several times. And load is seasonal: JEI warns that May through September is its busiest period, with typical delays of about two weeks at each stage, and asks authors to wait at least a week beyond the published estimate before chasing.
So: be prompt yourself, and patient with them. If you are planning around a deadline of your own, our guide to building a research project timeline covers how to budget for review.
What reviewers are actually asking for
Most comments on student papers cluster into five recurring asks, and knowing them in advance means writing a paper that avoids them.
"What is new here?" The originality question. Reviewers want an explicit statement of what your work adds. This is the same demand the AP Research rubric makes at its top band, which asks for explicit links to scholarly work that identify a research gap. Our guide to what counts as original research covers how to answer it.
"Why this method?" Not what you did — why it was right for this question, and what it cannot show.
"Your conclusions overreach." Extremely common. The fix is almost always to narrow the claim rather than add evidence.
"The limitations are thin." A real limitations section names what would change your conclusion, not just that your sample was small.
"Attribution is unclear." Where the reader cannot tell which claims are yours and which are the literature's. Journals take this seriously — OJSS states that papers found to include plagiarism or AI-generated writing will not be considered, so keep your own voice and your sources plainly separate.
One last thing about authorship
If a supervisor or programme mentor becomes involved during revision, settle the authorship question before resubmission rather than after.
The ICMJE criteria are clear that authorship requires all four of substantial contribution, drafting or critical revision, final approval, and accountability — and that general supervision of a research group does not qualify someone for authorship on its own. Contributors who fall short should be acknowledged, with their permission, not added to the byline.
Our guide to student research paper authorship covers the conversation, and what happens after you submit covers the rest of the process.
A reviewer report means a competent stranger read your work carefully enough to criticise it in detail. For a seventeen-year-old's paper, that is already an achievement — and the revision is the part where the paper actually gets good.
Frequently asked questions
What does revise and resubmit mean?
That the journal is interested but will not publish the paper as it stands. It is a good outcome, not a rejection. The Oxford Journal of Student Scholarship describes a revise-and-resubmit round as a normal step before final acceptance, and most student papers that get published go through at least one.
Do I have to make every change a reviewer asks for?
No, but you have to respond to every one. Journals that organise their reports well distinguish required changes from recommended ones — JEI's editor's letter labels them explicitly — and required changes are not optional. For recommended changes you may decline, provided you give a reason in the response letter.
Can I disagree with a reviewer?
Yes, and well-handled disagreement is a sign of a competent author. Say what you understand the reviewer to be asking, explain why you have taken a different view, and cite something — your data, the literature, or a constraint of your design. What does not work is ignoring the comment or restating your original claim more firmly.
How long should a response letter be?
As long as the number of comments requires, and no longer per comment. One numbered entry per reviewer point, each saying what you did and where to find it. Three or four sentences each is normal. Editors read these looking for completeness, not eloquence.
How long do I have to revise?
The decision email will say, and it is usually weeks rather than months. If you need longer, ask before the deadline rather than after it. Journals are generally accommodating about a requested extension and much less so about silence.
What if the reviewers contradict each other?
Say so, politely, and ask the editor to arbitrate — or pick the one you find better supported and explain your reasoning in the letter. Contradictory reports are common and are the editor's problem to resolve, not a trap for you. Do not try to satisfy both at once if they are genuinely incompatible.