Predatory Publishing Risks: An Institutional Checklist

Institutions usually see predatory publishing late, at promotion or thesis deposit, so a simple journal triage and a clear escalation path catch it earlier.

Checklist · 13 min read · Directive Publications Editorial Team

Build a simple triage

For any unfamiliar journal, check:

1. Peer-review description
2. Editorial board verifiability
3. Fee transparency
4. DOI resolution
5. Honest indexing claims

The list is short deliberately. It has to run in about ten minutes on a laptop, by a librarian, a supervisor or the author, without turning into an approval process that people route around. Five checks is roughly what fits inside an office-hours conversation.

Peer-review description. Read what the journal says about its own review and ask whether a stranger could tell what will happen to a manuscript. A usable description names the model — single-anonymous, double-anonymous, open — says how many reviewers a paper normally goes to, says who takes the decision, and gives a typical time to first decision in weeks. A useless one says that all submissions are peer reviewed by experts in the field and stops there. Absence is a finding. So is impossible speed: review time is governed by how long busy academics take to agree and then read, which no publisher controls, so a guaranteed decision within a few days is a promise about something the journal cannot deliver. The error this check prevents is specific and very common — seeing the phrase "peer reviewed" on a page and treating the phrase as evidence of the practice.

Editorial board verifiability. Check the board from the other end. Pick three names, ideally the ones with the most recognisable affiliations, since those are the names worth borrowing, and look for the role on the person's own institutional page, ORCID record or CV. Then write to one and ask politely what they last handled for the journal. Two things turn up. Names that belong to nobody, and names that belong to real researchers who never agreed to appear and in some cases have been asking to be removed for a long time. The shape of the board is informative too: dozens of names spanning unrelated disciplines, no identifiable editor-in-chief, or contact addresses that are all free webmail rather than institutional. What goes wrong when this check is skipped is that the board is the most persuasive thing on the site, and it is also the cheapest thing to fabricate.

Fee transparency. You should be able to write down the total cost before you submit: the amount, the currency, what triggers it, what it covers, whether waivers exist and how to apply, and what happens if you withdraw. A charge that appears only after acceptance is the classic pattern, because acceptance is the moment an author has the least room to walk away. Watch also for a submission or handling fee followed by a second charge later, and for an invoice whose amount does not match the published figure. The institutional version of this failure is a finance problem rather than an ethics one: a publication charge is paid from a grant, and afterwards nobody can produce a decision letter, a reviewer report, or even an archived fee page showing what was bought.

DOI resolution. Take a DOI from a recent article, put it after the doi.org resolver address, and see where it lands. It should arrive at that specific article's page on the publisher's site. Three failures matter: no DOI at all; strings that look like DOIs, appear on the PDF, and resolve to nothing; and DOIs that resolve to the journal homepage rather than the article. While you are there, look for a statement about long-term preservation, because an identifier only persists if somebody has arranged for the content to outlive the publisher. This is the check with the most direct institutional consequence, since repositories, funder reporting, reference managers and CV verification all run on identifiers. A paper with no resolving DOI is awkward to deposit, awkward to cite, and hard to prove.

Honest indexing claims. Verify indexing at the index, never on the journal's own page — an indexing claim is a line of text that costs nothing to write. Search the database itself for the journal title, and know what each kind of listing actually means: selective databases apply criteria and can remove titles, directories list journals that meet baseline transparency requirements, and being crawled by a general scholarly search engine is not indexing in any evaluative sense. The same scepticism applies to metrics. If a journal advertises a score under a name you do not recognise, treat it as marketing until you can find who calculates it and from what. Institutions are more exposed here than individuals, because indexing is precisely the claim that promotion criteria, scholarship rules and internal funding policies tend to lean on.

Combining the five is not a score out of five. One hard failure ends it — a board member who denies the role, a fee that only exists after acceptance, DOIs that resolve nowhere. The five together are for the ambiguous middle, and the middle is real. A new society journal in a narrow field may have a small board, no metric worth quoting, a plain website and a thin archive, and be entirely legitimate. The test is deception, not production values. The most reliable marker of an honest small journal is that it is candid about what it is not: no indexing claim it cannot support, no invented metric, a fee page that says plainly what it says.

Trainer resource: How to Identify a Predatory Journal

Process ideas that work

• Library “journal check” office hours
• Mandatory module before first submission
• Clear guidance in graduate handbooks
• Escalation path to the research integrity office

What these have in common is that none of them depends on a researcher deciding, unprompted, to be careful. Awareness campaigns reach the people who were already going to check. Structure reaches everybody else.

Library journal check office hours work because the barrier is usually not knowledge, it is not wanting to look naive by asking. Run a standing slot at the same time every week, no appointment needed, twenty minutes, and give a verdict inside the session with the checks written down so the author can show a supervisor what was found. Keep a running log of every title asked about, since the second enquiry about the same journal then takes two minutes. Advertise it where the money and the deadlines are, not only to authors: department administrators who raise purchase orders, grant officers, thesis coordinators, anyone who processes an invoice.

A mandatory module before first submission only stays mandatory if it is attached to something that already happens. Tie it to confirmation of candidature, to induction for new postdocs, or to the first request to pay a publication charge from institutional funds, so that compliance does not rely on enthusiasm. Keep it under an hour, and assess it by having people run the triage on a real journal page and write down what they checked, rather than with a quiz about definitions. Include one ambiguous case — a small, legitimate, unglamorous journal — or you will teach pattern matching on ugly websites instead of verification.

Guidance in graduate handbooks should read as procedure, not warning. Name the five checks, name the person or address to ask, and state what the institution will and will not pay for. Then state the consequences in advance. If publications in outlets that cannot be verified do not count toward a thesis-by-publication requirement, or toward promotion, that belongs in the handbook while a student can still act on it, not in an email at deposit. Vague cautions about being careful produce anxiety and no change in behaviour; a named checklist and a named contact produce both.

An escalation path to the research integrity office needs two doors, clearly distinguished. "Help me check this journal" goes to the library. "I think something has already gone wrong" goes to integrity. Publish a role and a monitored address rather than an individual who may move on. Say in plain words that asking about your own submission is not an admission of anything, and that the first response to a researcher who has already paid is help rather than a case file. If the path feels like reporting yourself for misconduct, nobody uses it, and you are back to finding out at promotion.

Watch for campus warning signs

• Sudden clusters of publications in unknown outlets
• Cold-email special issues with guaranteed acceptance
• APCs paid without documentation of review

These are institution-level signals, visible to a research office and invisible to any individual author. They are the reason this is a workflow problem: no single researcher can see a pattern that only exists across a department.

Sudden clusters of publications in unknown outlets show up in places you already look — an annual departmental output list, repository ingest, promotion dossiers, a CV. A cluster almost always has a shared cause: one lab, one visiting cohort, one supervisor's advice, or one incentive scheme that attaches a bonus or a graduation requirement to a publication count. Read it as a finding about the system before you read it as a finding about people, and check your own rules first. If promotion or bonus criteria count publications without asking where they appeared, the institution has created the demand it is now trying to police.

Cold-email special issues with guaranteed acceptance target mid-career researchers with a flattering invitation to guest edit, a request to bring a number of papers from their own network, and language about acceptance that arrives before anything has been read. What is being bought is legitimacy: a real name and a real address book. When a colleague forwards one, three questions settle it. Who at the parent journal appointed you, and can you find that person independently? Who selects reviewers and who takes the decision — you, or the office that sent the email? And what happens if you reject a paper submitted by someone in your own circle? Evasive answers, or a target number of papers, end the conversation.

APCs paid without documentation of review is the signal your finance system can catch, and finance is the only part of the institution that sees every payment. Add one requirement to the payment or reimbursement form: the acceptance letter, and a link to or copy of the journal's published fee page. The point is not to police researchers. It is that this is the last stage at which anything is recoverable — money still inside the institution can be stopped, and a paper not yet submitted can go elsewhere. It also produces, as a by-product, a list of every outlet your institution has paid, which is the closest thing you will get to a live map of your own exposure.

Related: Fake Peer Review · Special Issues

A worked example

A second-year PhD student forwards an invitation. It uses her name, praises a paper she published last year, offers rapid review, mentions a special issue, and offers a discount for submitting this month. Rather than telling her it looks suspicious, sit with her and run the triage on screen. It takes about ten minutes.

The peer-review page says manuscripts undergo rigorous peer review and promises a first decision within seventy-two hours. That is two findings, not one: no model, no reviewer count, nobody named as the decision-maker, plus a turnaround nobody can staff. The board lists dozens of names across half a dozen unrelated fields. Pick the three from the best-known universities; two have no mention of the journal anywhere on their own pages, and the third does not reply. The fee page says charges depend on article type and will be confirmed after acceptance — that sentence is the finding, and it is enough on its own. Take two recent articles: one carries no DOI, and the other's DOI resolves to the journal homepage rather than to the paper. The site lists several databases and a score with an impressive name; searching two of those databases for the journal title returns nothing, and the name of the score leads to no methodology.

Now the part that matters more than the verdict. Do not say "this is a predatory journal", which is an assertion that is hard to defend and easy to argue about. Say what failed, in order, with what you checked and where you checked it, and let the record speak. Then move straight to the next step, which is building a shortlist from her own reference list — the journals she already cites are the journals whose readers need the result. The invitation arrived because she had published something; the durable fix is not more suspicion of email, but having chosen the journal before anyone wrote to her.

If a paper is already published in one of these journals

Order matters here, and the first instruction is not to delete anything. Preserve the record: save the article page, the acceptance email, the invoice, and every message exchanged with the journal. These sites vanish, and a researcher who later has to explain a payment to a funder or a panel will need what can no longer be retrieved.

Then establish what actually exists. Does the DOI resolve? Is the article findable anywhere outside the journal's own site? Is there any preservation arrangement? The answers decide whether you are dealing with something that will persist in the scholarly record or with a page that may simply disappear, and the two need different handling.

Ask for withdrawal in writing, and be realistic about the response. In the meantime, one instruction to the author matters more than the others: do not submit the same manuscript elsewhere while it remains published. That is duplicate publication, it becomes the author's problem rather than the journal's, and it turns a bad decision into a finding against the researcher. If a withdrawal is confirmed in writing, keep the confirmation and disclose the history to the next editor at resubmission. Editors meet this situation more often than authors imagine, and the disclosure is what protects the author.

Everything after that is a judgement to make with the integrity office rather than alone: how the item is described on a CV, what the funder or the repository is told, whether a thesis needs restructuring. Handle the person carefully — in most cases you are looking at someone who was given no guidance and acted in good faith, and the way the first case is handled decides whether the second one ever reaches you voluntarily. Then treat it as a system finding. One researcher who submitted to an unchecked outlet usually means the checking step does not exist for anybody.

Bottom line

Predatory risk is a training and workflow problem — not only an individual mistake. Use Directive Academy in workshops.

The test of whether you have fixed it is not awareness, it is timing. Ask when your institution most recently learned that someone had published in an outlet it could not verify, and what the trigger was. If the answer is a promotion panel, a thesis deposit or a funder query, detection is still happening at the end of the process, where nothing can be changed. If the answer is a librarian's office hour, a question raised in a module, or a query on a payment form, it has moved to the front — and that is the entire objective.

More: Directive Academy · For Authors