A Reviewer's Main Responsibility Is To… (Answer Explained)
If you are working through a research ethics or peer review training module, you have probably met the question "A reviewer's main responsibility is to…" followed by four options. Here is the answer, why the other options are wrong, and what the answer means when you sit down to review real work.
The question and the four options
The question appears in responsible conduct of research (RCR) and peer review training, and in the flashcard sets students build from it. The wording changes slightly between versions, but the options usually look like this:
| Option | Correct? | Why |
|---|---|---|
| Avoid criticizing the work of other researchers | No | Critical evaluation is the whole point of a review. A reviewer who avoids criticism is not doing the job. |
| Be professional and prompt (or "behave professionally") | Yes | Every other duty, from fairness to confidentiality to meeting the deadline, follows from professional conduct. |
| Be friendly | No | Courtesy matters, but friendliness is a tone, not a responsibility. A friendly review that misses a fatal flaw has failed. |
| Contact the authors when the review is completed | No | Reviewers send their report to the editor. Direct contact with authors is something reviewers are told to avoid. |
Why "professional and prompt" is the right answer
Peer review runs on trust. Authors hand over unpublished work, editors rely on the reviewer's judgment, and readers assume that what gets published was checked by someone competent. None of that works unless the reviewer behaves like a professional: honest about what they know, fair to people they may compete with, discreet with what they read, and on time.
The "prompt" half is not a minor detail. A late review holds up a paper that someone's degree, grant, or job may depend on. The Council of Science Editors lists unreasonable delay next to breaching confidentiality as an example of reviewer impropriety, and says a reviewer who cannot meet the deadline should decline quickly or ask the editor whether more time is possible.
What professional conduct means: six duties
"Be professional" is easy to tick on a quiz and harder to practice. Journal and editor guidelines break it down into the same handful of duties.
1. Review only what you are competent to judge
You do not need to be an expert in every part of a manuscript, but you need enough expertise to give an authoritative opinion. If part of the work is outside your field, such as the statistics, say so to the editor so they can find someone else for that part. If most of it is outside your field, decline.
2. Keep the manuscript confidential
A manuscript under review is a privileged document. Do not share it, discuss it with colleagues, keep copies, or use its ideas or data in your own work. If you want a colleague or student to help, ask the editor first.
3. Be impartial and declare conflicts of interest
Judge the work on its merit, originality, and clarity, not on who wrote it or where they are from. If you have a personal, financial, or competitive relationship with the authors that could affect your judgment, or could look as if it does, tell the editor before you start. When in doubt, ask.
4. Give constructive, evidence-based criticism
Point out what works as well as what does not, explain your reasoning, and tell the authors what would fix each problem. The purpose is to improve the paper and inform the editor's decision, not to show how many flaws you can find. Personal remarks about the authors have no place in a review.
5. Be on time
Answer the invitation quickly, even if the answer is no. If you accept, deliver by the deadline. If something changes, tell the editor straight away so the manuscript can be reassigned.
6. Flag ethical concerns to the editor
If you suspect plagiarism, duplicate publication, manipulated data or images, or problems with how human or animal subjects were treated, report it confidentially to the editor. It is not your job to investigate or to accuse the authors yourself.
Who a reviewer is responsible to
The Council of Science Editors groups reviewer responsibilities by the three parties who depend on the review.
| Responsible to | What the reviewer owes them |
|---|---|
| Authors | Written, unbiased, constructive feedback delivered on time; no personal criticism; confidentiality. |
| Editors | A quick reply to the invitation; disclosure of conflicts; a fair and informative critique; a recommendation if asked; notice of any ethical concern; no direct contact with authors. |
| Readers | A check that methods are described well enough to judge and repeat, and that relevant earlier work is cited. |
One point surprises new reviewers: the reviewer does not decide whether the paper is published. As the InterPore Journal's reviewer duties put it, the reviewer's role is to help the editor evaluate the manuscript and to help the authors improve it. The decision belongs to the editor.
What a reviewer should not do
- Delay the review without good reason.
- Criticize a competitor's work unfairly, or misstate facts.
- Use ideas, data, or text from the manuscript before it is published.
- Ask the authors to cite the reviewer's own papers when those papers are not needed.
- Hide a conflict of interest.
- Bring in another person to help without the editor's knowledge.
- Write comments about the authors instead of about the work.
The European Association of Science Editors publishes a short list of dos and don'ts that covers the same ground and is worth keeping open while you write your first report.
Professional conduct and AI tools
Confidentiality now includes what you paste into a chatbot. Many journals tell reviewers not to upload a manuscript, or any part of it, into a generative AI tool, because doing so can breach the authors' confidentiality and rights. Policies differ on the review report itself: IOP Publishing, for example, allows reviewers to use such tools only to improve the writing of their own report and not to analyze or evaluate the manuscript, while other publishers do not allow the report to be uploaded at all. Read the policy of the journal you are reviewing for before you use any tool.
The same rule applies to every kind of reviewer
The quiz question is about peer review, but the answer holds for anyone whose opinion other people rely on.
- Product reviewers owe readers an honest account based on real use, and a clear statement if the product was free or the link earns a commission.
- Code reviewers owe teammates specific comments about the code, not the coder, without leaving the pull request waiting for days.
- Book and film reviewers owe the audience a fair description of what the work tries to do before judging whether it succeeds.
- Performance reviewers at work owe employees evidence, examples, and confidentiality.
In each case the job is the same: know the subject, be fair, say what you found and why, keep private things private, and deliver when you said you would. If you are about to write your first journal review, our step-by-step guide to reviewing a journal paper walks through the process from invitation to recommendation.
Frequently asked questions
What is a reviewer's main responsibility?
A reviewer's main responsibility is to be professional and prompt: to give a fair, competent, confidential and constructive evaluation of the work, and to deliver it on time.
Should a reviewer contact the authors after finishing the review?
No. The reviewer sends the report to the editor, who passes comments to the authors. Journals tell reviewers not to contact authors directly.
Is a reviewer supposed to avoid criticizing other researchers' work?
No. Critical evaluation is the purpose of a review. The criticism should be about the work, backed by reasons, and written so the authors can act on it.
What should a reviewer do about a conflict of interest?
Tell the editor before starting. If the conflict could affect the reviewer's judgment, or look as if it does, the reviewer should decline.
Does the reviewer decide whether a paper is published?
No. The reviewer advises. The editor makes the decision, usually after reading reports from two or more reviewers.