An associate professor accepts an invitation to review a manuscript very close to her group’s own work. She passes it to a PhD student, who writes the report; she tweaks a few sentences and submits it under her name. A month later the student proposes a “new” experiment at a lab meeting that looks remarkably like the manuscript’s methods. Nobody meant harm, yet in a chain of small decisions three basic principles of peer review were broken: an undeclared conflict of interest, unauthorised sharing of a manuscript, and reuse of unpublished ideas.
COPE’s ethical guidelines for peer reviewers are short and readable. This article turns them into the situations new reviewers actually run into.
The manuscript belongs to someone else
When you accept, you are holding unpublished work that may represent years of effort. Confidentiality means:
- Do not share the manuscript, even with a trusted colleague, unless the editor agrees.
- Do not use its data, methods or ideas in your own work before it is published.
- Do not contact the authors, even if you are fairly sure who they are.
- Do not upload it to outside tools (machine translation, AI assistants, personal plagiarism checkers) unless the journal permits it.
- Delete or destroy your copy when you are done, unless the journal says otherwise.
When you should decline
Conflicts of interest are not only financial. Decline, or tell the editor straight away, if:
| Situation | What to do |
|---|---|
| You co-authored with one of the authors in recent years | Decline, or declare it and let the editor decide |
| You work in the same department or institute | Decline |
| You are about to submit directly competing work | Decline |
| A supervisory, family or close personal relationship, or a known personal dispute | Decline |
| You have a financial interest in the outcome | Decline |
| You already reviewed this manuscript for another journal | Tell the editor; many will still want you, but they need to know |
| You lack expertise for part of it, such as the statistics | Accept, but state clearly which parts you cannot assess |
One more legitimate reason to decline: no time. Saying no within two days is far kinder than accepting and being six weeks late while the authors wait.
Co-reviewing with a PhD student, done properly
Reviewing alongside a student is excellent training, and many journals welcome it. What goes wrong is doing it quietly. The right steps:
- Ask the editor before sharing the manuscript. Many systems have a field for declaring a co-reviewer.
- Name the co-reviewer in your comments to the editor so they receive credit.
- Explain the confidentiality duty to the student exactly as it applies to you.
- Keep final responsibility: read the whole manuscript yourself rather than just polishing the student’s report.
Ghost reviewing, where the person who actually wrote the report goes unacknowledged, shortchanges junior researchers and leaves editors unaware of who really assessed the paper.
Writing the report: fair and impersonal
- Comment on the work, not the people. “This analysis omits confounder X” rather than “The authors clearly do not understand statistics.”
- Do not judge on nationality, first language or institutional prestige. Note language problems briefly and do not let them drive your recommendation.
- Do not request citations to your own work without a specific reason.
- Separate essential revisions from optional suggestions so authors know what matters most.
- Confidential comments to the editor are not the place for criticism you would not make to the authors; they are for integrity concerns or sensitive information.
If you suspect misconduct in the manuscript
A figure looks reused, a paragraph seems familiar, or numbers do not add up. Your job is to tell the editor confidentially, describing specifically what you saw. Do not investigate, contact the authors or discuss it with others. Phrase it as an observation: “Panels 2C and 4A appear to share a background region; I suggest the editorial office request the original images,” not a verdict such as “the authors committed fraud.”
Open review, blind review and revealing your identity
Under single- or double-blind review, do not sign your report and do not reveal that you reviewed the paper, even after publication, unless the journal allows it. Under open review your report may be published with the article, so write as if it will be read publicly (which is a good standard under any model). Listing your reviewing activity on a research profile is fine as long as you disclose nothing about the manuscript itself.
A checklist before you click “accept invitation”
- Do I have any conflict of interest?
- Do I have the expertise for most of the manuscript, and which parts will I say I cannot assess?
- Can I realistically deliver on time?
- If I want to involve a student, have I asked the editor?
- Will I use any outside tool with the manuscript, and does the journal allow it?
If you have never reviewed, start by reading one journal’s reviewer guidelines in your field and registering as a reviewer with two or three journals you read regularly. Do the first couple of reviews well and editors will remember your name, which is a credible route to visibility that few early-career researchers use.
Câu hỏi thường gặp
Can a peer reviewer share a manuscript with a PhD student?
Only with the editor’s permission. Many journals welcome declared co-reviewing, and the co-reviewer should be named in the comments to the editor.
When should I decline a peer review invitation?
Decline when you have a conflict of interest such as recent collaboration, the same institution, competing work, a personal relationship or a financial stake, or when you cannot deliver on time.
Can I use ideas from a manuscript I am reviewing?
No. Data, methods and ideas in an unpublished manuscript cannot be used in your own work. Once the paper is published, you can cite it like any other source.
What should a reviewer do if they suspect plagiarism?
Inform the editor confidentially with specific observations, without investigating yourself or contacting the authors. Verification is the editorial office’s job.
Can I list peer review activity on my CV?
Yes, you can list the journals and number of reviews, or use a review recognition service. Do not reveal titles or content of manuscripts reviewed under blind review.
