For peer reviewers
Review, and be credited for it
Peer review fails when it is slow, thankless and invisible. Two thousand one hundred scholars review with us under an arrangement that fixes all three: a four-week window, an honorarium per completed report, and a reviewing record pushed to your ORCID.
The reviewer college
The college is the standing body of reviewers across all six journals. Membership is by application or by invitation from a handling editor, and it carries no obligation to accept any particular request — you decline by clicking decline, and declining costs you nothing.
What it does carry is a record. Every completed review is logged against your ORCID with the journal and the month, without ever naming the manuscript, so a reviewing history becomes something you can put in a promotion file.
What we ask
- Four weeks
- From acceptance of the invitation to the report. If you need longer, say so at the invitation stage and we will usually agree; what we cannot do is find out in week five.
- A structured report
- Against the template below rather than free-form prose. Structure is what makes two reports comparable to an editor and useful to an author.
- A recommendation, and reasons for it
- Accept, minor revision, major revision or decline — with the reasoning attached to specific claims, sections or analyses rather than to the paper as a whole.
- Decline promptly if it is not yours
- Outside your competence, too close to your own work, or a conflict — decline in a day rather than in a fortnight. It is the single most useful thing a reviewer can do.
Credit & honoraria
Reviewing is skilled professional labour and we pay for it. This is unusual and we would rather it were not.
- An honorarium per completed review
- Paid on receipt of the report, not on the editorial decision. A report that recommends decline is worth exactly what one recommending acceptance is worth.
- ORCID-recorded
- Journal and month, pushed to your ORCID record automatically. The manuscript is never named.
- Named on the article, if you want to be
- Where both you and the author opt into open review, your report is published alongside the article under your name.
- Annual acknowledgement
- Every journal publishes its reviewer list once a year. You can opt out of appearing on it.
How to review
- 1Read the invitation, then the abstractDecide on competence and conflict before you read further. If either is a problem, decline now.
- 2Read the whole paper once, without notesYou are establishing what the paper is trying to do. Notes taken in the first pass tend to be about the paper you would have written.
- 3Read it again, against its own claimsDoes the method answer the question asked? Do the results support the conclusions drawn? Is anything claimed that is not shown?
- 4Separate the fatal from the fixableAn editor needs to know which of your points are conditions of publication and which are suggestions. Say which is which explicitly.
- 5Write to the author, not about themThe report is read by a person who has spent two years on this. Be direct about the work and courteous about the author; the two are entirely compatible.
The report template
Every report on this platform is structured in the same six parts.
- Summary
- Two or three sentences on what the paper claims and does. If you cannot write this, that is itself a finding worth reporting.
- Significance
- Would this change what someone in the field does or believes? Say so plainly, in either direction.
- Major points
- Numbered. Each one a condition of publication, tied to a specific section, figure or analysis.
- Minor points
- Numbered. Things that should be fixed but do not decide the outcome.
- Data, code and reporting
- Are the availability statements adequate? Was the right reporting guideline followed? Could you reproduce the analysis from what is provided?
- Confidential comments to the editor
- Anything you would not say to the author, including suspicions of misconduct. These are never shown to the author.
Reviewer ethics
- Confidentiality
- A manuscript under review is confidential. Do not share it, cite it, or use its findings before publication — including as a reason to hurry your own similar work.
- Conflicts
- Declare at invitation: same institution, recent collaboration, a competing manuscript in preparation, or any personal relationship. Declaring is not disqualifying by itself; the editor decides.
- Delegation
- You may not pass a review to a student or colleague without telling us. You may co-review with one, named, and we encourage it.
- Generative AI
- Do not paste a manuscript into a public AI tool — it is a confidentiality breach regardless of what the tool does with it. AI assistance in drafting your own prose is fine and should be disclosed.
- Suspected misconduct
- Report it in the confidential comments rather than acting on it. The editor and the ethics team handle it under our published policy; you are not asked to investigate.
Apply to review
Tell us your field, your methods and your ORCID. We match on both subject and method — a paper whose statistics are the contested part needs a statistician, not another subject specialist who will defer to them.
