Question practices rather than applying a simple label
The term “predatory” is commonly applied to outlets that misrepresent or fail to provide expected scholarly publishing services. It remains contested because journals vary in resources, business model and quality. A fair investigation describes observable conduct instead of attaching assumptions to individuals, countries or open-access publishing as a whole.
Charging authors is not evidence of poor quality. Many legitimate journals use article processing charges, while others rely on subscriptions or institutional support. Ask whether costs are disclosed before submission, whether waivers and timing are clear, and whether the stated editorial and preservation services are actually provided.
Use warning signs to trigger checks, not conclusions
Potential signals include implausibly fast acceptance promises, aggressive invitations unrelated to the recipient's field, a name copied from an established journal, unreachable contacts and editorial members who cannot be verified. Invented metrics, false indexing claims and fees revealed late also warrant investigation. Awkward language or a modest website on its own does not.
Read a sample of published articles. Do they fit the scope? Are received, accepted and published dates presented plausibly? Can you find corrections, retractions and competing-interest statements? Resolve several DOIs and inspect the continuity of older issues. An isolated technical fault can occur anywhere; a pattern of misleading claims is more informative.
Check each prestige claim at its stated source
When a journal claims inclusion in DOAJ, Web of Science, Scopus, PubMed or a subject index, search the title or ISSN in that service's current official record. A logo on the journal site is not evidence. Confirm that the listing matches the exact title, publisher and relevant years rather than a similarly named publication.
Open institutional profiles for editorial-board members. Names can be listed without meaningful involvement or consent. Compare publisher address, domain and contact route, but do not treat a post-box address alone as proof. Look for accessible policies on peer review, plagiarism, data, authorship, competing interests, complaints, corrections, retractions and long-term preservation.
Why a blacklist can produce false confidence
Lists can age, hide criteria, confuse titles or fail to record improvements. Conversely, a newly deceptive operation may not yet appear. Treat any list as a lead and investigate the current journal yourself. Inclusion in a trusted directory is useful evidence, but not a substitute for evaluating the particular article and publication period.
Account for differences in power and language. A small scholarly community may have an unsophisticated site without misrepresenting peer review. Regional subject matter can be underrepresented in international indexes. Evaluate transparent, verifiable practice rather than prestige, nation or polished English-language marketing.
Make a decision that another person can audit
Before submitting a manuscript, also inspect copyright transfer, licence, fee timing and withdrawal terms. Discuss borderline cases with a librarian, learned society or supervisor. Do not send confidential manuscript data to an unverified contact. Preserve the invitation, relevant web statements and your dated findings.
When assessing an article already published there, continue to evaluate its method and evidence. A reputable journal does not make every paper strong. A troubling venue does not make every observation logically false, but it materially raises the verification burden. Look for underlying data, independent replication and expert context. The guide to identifying scholarly sources provides the general article-level assessment.
Use a traffic light without pretending it is a score
Green means identity, policies and claims can be substantiated; the paper still needs critical reading. Amber means consequential details are missing or inconsistent and require clarification before submission or central citation. Red means several core claims are demonstrably misleading or essential publishing services cannot be evidenced. Write the reasons beside the colour.
Do not turn the matrix into an automatic point system. The importance of a feature depends on context: a fabricated editorial board is more serious than a broken link. Date the conclusion and revisit it if ownership, policies or indexing change. A neutral description such as “the claimed index listing could not be verified on 4 September 2026” is stronger than an unsupported accusation.