Recognising Peer Review in Promotion and Tenure

Reviewing holds research quality together, yet most promotion files reduce it to one line nobody can check. Committees can do better without breaching confidentiality.

Explainer · 10 min read · Directive Publications Editorial Team

Why reviewing disappears from promotion files

A candidate may spend days each year reviewing manuscripts and grant applications, and the promotion file shows one line: "Reviewer for several journals." The committee cannot read the reports, check the claim, or tell a careful methods critique from a two-sentence recommendation to accept.

So reviewing is usually either ignored or counted, and neither works. Ignoring it tells staff that the process every department relies on earns no credit. Counting it rewards whoever accepts the most invitations, including invitations from journals that should never have sent them.

The Hong Kong Principles for assessing researchers, developed as part of the 6th World Conference on Research Integrity, include a principle to recognise essential tasks such as peer review and mentoring. They state that "the quality of peer-review contributions to journals and funders should also be part of assessments for promotion and tenure." The first commitment of the Agreement on Reforming Research Assessment names peer review among the practices assessment should be able to recognise, and adds that no individual researcher should be expected to contribute to every activity at once.

For a medical school the stakes are sharper. Clinical academics on a joint contract already split their time between service, teaching and research, and reviewing is the part of the research half that leaves no trace in a publication list. Statistical and methods reviewing for clinical journals, and health-funder grant panels, are among the most demanding forms of the work and the least visible in a file.

Decide where reviewing sits in your criteria

Before you ask candidates for evidence, decide where that evidence goes. Institutions place peer review in different categories, and where the criteria are silent, each committee improvises and candidates are treated inconsistently. There is no single correct answer, but there should be a written one. Classify each activity by what it demonstrates, and count it once.

Tell candidates which category you use before they assemble a dossier. Moving reviewing to a different category at the committee stage penalises the people who followed the old guidance.

How to verify reviewing that was anonymous

When review is single- or double-anonymised, the committee cannot see the report and the candidate should not show it. Verification has to confirm that a review happened, for whom and when, without disclosing the manuscript. Three sources do that.

ORCID peer review entries. ORCID's help pages on the peer review section say reviews "can only be added to your ORCID record by a trusted organization using Member API credentials, with your explicit permission — you cannot manually add them to your record yourself." Double-anonymised entries carry minimum information; open reviews can point at the report itself. ORCID also names the source of each item, meaning whoever added it, and that matters twice over. ORCID directs researchers to record serving as a reviewer as a Service affiliation, which they add themselves, so treat a self-added entry as a claim rather than a verification. And when a review reaches ORCID through a recognition service rather than the journal, the source line shows that service: a Web of Science-sourced item says Web of Science verified the review, not that the journal asserted it.

Reviewer recognition services. A Web of Science Researcher Profile, Clarivate's successor to Publons since August 2022, holds verified records of reviews, editor records and publications in one place. Reviews for partnered journals are added through the submission system when the reviewer opts in; reviews for other journals can still be verified by sending Clarivate a review receipt, typically the journal's email confirming a completed review, carrying the journal's full title and the date. By default a public profile shows only the journal or conference name and the number of verified reviews, and no manuscript information before publication. The profile can be exported as a CV, which Clarivate says researchers can use "in funding, promotion, job, or green card applications". Do not confuse it with the Reviewer Recognition Service, which Clarivate sells to publishers.

Confirmation from the journal. Where neither record exists, a candidate can ask the journal office for a short note confirming the number of completed reviews over a stated period, with no manuscript details. For a clinical journal, ask the note to say whether the reviewing was general, statistical or methodological, since those roles carry different workloads.

Your guidance must also say what not to include, because COPE's Ethical Guidelines for Peer Reviewers ask reviewers not to reveal details of a manuscript after peer review without permission from both the author and the journal. If your guidance is silent, a conscientious candidate may breach confidentiality trying to prove the work, and an unbriefed committee may ask them to.

None of these records says anything about the quality of a review. They establish that reviewing happened: the floor, not the assessment.

Treat editorial roles on their own terms

Editorial titles vary so much between journals that the title alone tells you little. A handling editor may manage a steady stream of manuscripts; a named board member may be consulted rarely. Ask for a description of the work rather than the title:

• What did you do: handle manuscripts, make or recommend decisions, commission content, or advise occasionally?
• Roughly how much work did it involve, over what period?
• How were you appointed, and when does the term end?

A board listing with no handled manuscripts is an affiliation, not a contribution. Guest editing needs the same treatment: what matters is whether the candidate ran a real editorial process, which our post on special issues and guest editing describes.

Apply the same source test to titles that you apply to reviewing claims. ORCID now offers Editorial Service as an affiliation type, covering editorial board member, associate editor and editor-in-chief, and a researcher can add one manually; ORCID records whether the researcher or a trusted organisation added it. A title the candidate typed in themselves is not the journal confirming it, so where a title carries weight, ask for the appointment letter or check the masthead.

Check the journal as well as the role. A title at a journal that fails basic legitimacy checks should earn no credit, and a pattern of such titles is worth a conversation with the candidate. The editorial board checks in how institutions can spot predatory publishing risks work just as well when the name on the board belongs to your own staff member.

Weigh quality, not volume

A number is the easiest part of a reviewing record to present and the least informative. A candidate who writes many brief reports has not necessarily contributed more than one who writes fewer detailed reports for demanding journals and funders. If your criteria reward a count, you have told staff to accept more invitations. Ask for evidence a committee can judge instead:

1. A short narrative. A few paragraphs on the reviewing the candidate does: the expertise it draws on, any specialised role such as statistical or methods review, and any sustained reviewing relationship with a journal or funder. In clinical fields the specialism is the whole point: a statistical reviewer sent trial manuscripts after the subject reviewers report, a reviewer called on for reporting-guideline compliance, or one who checks a trial against its registered protocol is doing work the journal cannot easily source elsewhere. Ask candidates to name that role rather than describe their reviewing generically.
2. Readable examples, where they exist. Some journals publish reviewer reports, sometimes signed and registered with their own DOIs through Crossref. These are the only reviews a committee can read directly; our guide to transparent peer review explains that model.
3. Comment from editors or panel chairs. A letter describing how useful and reliable a candidate's reviews have been tells you more than any total.
4. Evidence of developing other reviewers. COPE's guidelines say supervisors who want to involve students in a review must ask the editor's permission, and that a student who performs a review under supervision should be acknowledged as the reviewer of record. Credit mentors who co-review that way, and not supervisors who pass work to a trainee without the journal's knowledge.

Be careful not to penalise people for receiving fewer invitations. Early career researchers, part-time staff and those returning from leave may review less for reasons unrelated to expertise.

What candidates should put in the dossier

Give candidates a template, so every file arrives in comparable shape:

And a checklist before submission:

1. Reviewing described in a narrative, not only listed
2. ORCID iD included, peer review entries visible to everyone
3. Confirmations attached where no verified record exists
4. No unpublished reports, titles or author names from confidential reviews
5. Editorial roles described by work done, with dates
6. Published reports linked by DOI, where they exist
7. Each activity listed under one category only

Build it into your criteria

For a research office, the first practical step is one paragraph of policy that committees and candidates both read:

Then brief committee chairs on the verification sources above, so nobody asks a candidate for a confidential report, and check the paragraph against the rest of the handbook. If your institution runs a separate clinical-academic or clinician-educator track, write it into that track's criteria too, and say which category reviewing lands in there. A clinical academic whose research time is a named fraction of a joint contract is the person most likely to be told that reviewing is admirable but does not count.

Reviewing is one of the few scholarly contributions designed to be invisible. Recognising it fairly means building a way to see it without breaking the confidentiality that makes it work. When your committees want a concrete reference for what a journal asks of its reviewers, Directive Publications sets out its peer review process publicly.