Setting the agenda in research Comment ILLUSTRATION BY DAVID PARKINS DAVID BY ILLUSTRATION promise was doubtful and its validity unlikely to have been vetted. Predatory journals: no Predatory journals are a global threat. They accept articles for publication — along with authors’ fees — without performing promised definition, no defence quality checks for issues such as plagiarism or ethical approval. Naive readers are not the only victims. Many researchers have been duped Agnes Grudniewicz, David Moher, Kelly D. Cobey and 32 co-authors into submitting to predatory journals, in which their work can be overlooked. One study that focused on 46,000 researchers based in Italy Leading scholars and hen ‘Jane’ turned to alternative found that about 5% of them published in such medicine, she had already outlets1. A separate analysis suggests preda- publishers from ten exhausted radiotherapy, chemo- tory publishers collect millions of dollars in countries have agreed a therapy and other standard publication fees that are ultimately paid out definition of predatory treatments for breast cancer. Her by funders such as the US National Institutes Walternative-medicine practitioner shared an of Health (NIH)2. publishing that can protect article about a therapy involving vitamin infu- One barrier to combating predatory pub- scholarship. It took 12 hours sions. To her and her practitioner, it seemed lishing is, in our view, the lack of an agreed to be authentic grounds for hope. But when definition. By analogy, consider the historical of discussion, 18 questions Jane showed the article to her son-in-law (one criteria for deciding whether an abnormal and 3 rounds to reach. of the authors of this Comment), he realized it bulge in the aorta, the largest artery in the body, came from a predatory journal — meaning its could be deemed an aneurysm — a dangerous 210 | Nature | Vol 576 | 12 December 2019 ©2019 Spri nger Nature Li mited. All ri ghts reserved. ©2019 Spri nger Nature Li mited. All ri ghts reserved. condition. One accepted definition was based NO LIST TO RULE THEM ALL on population norms, another on the size of Assessments of which journals are likely to be predatory or legitimate do not tally, the bulge relative to the aorta and a third on an and titles can appear in both categories. There is no way to know which journals were considered for a list but left o, or which were not considered. absolute measure of aorta width. Prevalence varied fourfold depending on the definition Suspected Legitimate predatory journals journals used. This complicated efforts to assess risk and Cabells ‘predatory’ Cabells ‘verified’ interventions, and created uncertainty about 3 § who should be offered a high-risk operation . Beall’s* 10,406 10,077 DOAJ Everyone agrees that predatory publishers sow confusion, promote shoddy scholarship 0 979 and waste resources. What is needed is consen- 228 sus on a definition of predatory journals. This would provide a reference point for research 1,135 0 0 11,306 journals into their prevalence and influence, and would help in crafting coherent interventions. 0 To hammer out such a consensus and to 0 map solutions, we and others met in Ottawa, 31 1 6 Canada, over two days in April this year. The 43 participants hailed from 10 countries and 34 represented publishing societies, research Beall’s list highlighted the The DOAJ relies mainly funders, researchers, policymakers, academic issue of predatory journals, on information from but faced criticism over publishers. It regularly institutions, libraries and patient partners transparency and legal purges titles that do (that is, patients and caregivers who proac- threats from listed titles. It Some journals deemed legitimate by not meet quality ceased operation in 2017. the DOAJ were deemed predatory criteria. tively engage in research). Our focus was the by Beall’s and/or Cabells lists. biomedical sciences, but our recommenda- tions should apply broadly. *Informally assessed by University of Colorado Denver librarian Jerey Beall in ~2008–17; Pay-to-access lists from Cabells, a scholarly analytics company; §The Directory of Open Access Journals, a community-curated list requiring journal best practices such as peer review and statements on author fees and licensing. Here we put forward our definition. We 5. REF. FROM ADAPTED SOURCE: describe what it took to achieve consensus and A journal’s membership of agencies such as CVs. We discussed replacing the term entirely how we’ll move forward. COPE (the Committee on Publication Ethics), with language that recognizes nuances in curated indexes such as Web of Science, or publishers’ quality and motivation. Alterna- The definition being listed in the Directory of Open Access tives considered included ‘dark’, ‘deceptive’, The consensus definition reached was: Journals (DOAJ) is insufficient to guarantee ‘illegitimate’ and ‘acting in bad faith’. Ulti- “Predatory journals and publishers are entities quality. Predatory journals have found ways mately, we concluded that the term ‘preda- that prioritize self-interest at the expense of to penetrate these lists, and new journals have tory’ has become recognized in the scholarly scholarship and are characterized by false or to publish for at least a year before they can community. Implementation science suggests misleading information, deviation from best apply for indexing. that introducing new nomenclature would take editorial and publication practices, a lack of A scoping review comparing publications considerable resources, which we felt could be transparency, and/or the use of aggressive and about predatory journals found that their better put towards combating predatory pub- indiscriminate solicitation practices.” characterizations sometimes overlapped, lishing directly. So we recommend keeping the sometimes did not and sometimes directly word ‘predatory’ while noting its limitations. Road to consensus conflicted7. These inconsistencies suggest Since the term ‘predatory publishers’ was that crafting a practical definition would Details matter coined in 2010, hundreds of scholarly articles, require building consensus across researchers, Predatory journals are driven by self-interest, including 38 research papers, have been written publishers, research institutions and the usually financial, at the expense of scholarship. warning about them. Scientific societies and broader public. They are characterized by the following: publishers (including Springer Nature) have Participants in our summit completed a helped to establish the ‘Think. Check. Submit.’ three-round modified Delphi survey (a struc- False or misleading information. This applies campaign to guide authors. But it is not enough. tured technique to elicit input, offer feedback to how the publisher presents itself. A preda- More than 90 checklists exist to help identify and build consensus) that included 18 ques- tory journal’s website or e-mails often present predatory journals using characteristics such tions and 28 sub-questions. There were also contradictory statements, fake impact factors, as sloppy presentation or titles that include 12 hours of discussion, followed by 2 further incorrect addresses, misrepresentations of words such as ‘international’. This is an over- rounds of feedback and revision. the editorial board, false claims of indexing or whelming number for authors. Only three Crafting a consensus definition was hard. membership of associations and misleading of the lists were developed using research Even reaching agreement on the use of ‘preda- claims about the rigour of peer review. evidence4. Paywalled lists of quality journals tory’ was a challenge. Part of the group wanted and predatory journals show that there is an a term that acknowledges that some authors Deviation from best editorial and publica- appetite for clear, authoritative guidance. But turn to these outlets fully aware of their low tion practices. Standards here have been set these lists are inconsistent and sometimes quality; these scholars willingly pay to pub- out in the joint statement on Principles of out of reach5,6 (see ‘No list to rule them all’). lish in predatory journals to add a line to their Transparency and Best Practice in Scholarly Nature | Vol 576 | 12 December 2019 | 211 ©2019 Spri nger Nature Li mited. All ri ghts reserved. ©2019 Spri nger Nature Li mited. All ri ghts reserved. Comment Publishing (see go.nature.com/35mq7mj), of peer review, even though negligent peer problem, tracking numbers of publications issued by the DOAJ, the Open Access Scholarly review is often a prominent feature of pred- in predatory journals by discipline and geog- Publishers Association, COPE and the World atory journals. We are not saying that peer raphy. We will work with funders, institutions, Association of Medical Editors. Examples review is unimportant, only that it is cur- patients and other stakeholders to iteratively of substandard practice include not having rently impossible to assess. Unfortunately, develop resources to assess journal quality. We a retraction policy, requesting a transfer of many legitimate journals fail to make their are seeking funding to create and test a digital copyright when publishing an open-access peer-review processes sufficiently transpar- tool to achieve these goals. article and not specifying a Creative Com- ent, for instance by sharing peer reviewers’ Efforts to counter predatory publishing mons licence in an open-access journal. These comments and other data. At the moment, need to be constant and