June 21, 2016

Debasing the Currency of Science – The Growing Menace of Predatory Open Access Journals

Introduction

‘Call for papers and editors OA Journal of Desperately Seeking Your Money - (Gold OA) ISO 9001:2008 - Certified ISSN 2277-3754 - Scopus ISI Thompson Reuters Indexed - High Visibility - Impact Factor 3.7’

Such ‘announcements’ are part of our daily email landscape, even when we have efficient spam filters. The corresponding websites claim to have ‘rapid publication,’ ‘top innovation,’ ‘thirst for excellence,’ accompanied by an assortment of official-sounding labels, logos, and abbreviations. Most of us recognize such an organization for what it is: a pop-up website whose purpose is to lure us into sending manuscripts, which will be accepted, on condition of payment of page charges. They are predatory journals.

Beyond the wry smile or annoyed ‘delete’ they may elicit, these ‘journals’ have created problems so serious that we believe they threaten the very existence of science, more so than at any time since it began to emerge in 17th century Europe — and this peril has nothing to do with lack of funding. Rather, the very currency of science is being rapidly debased by thousands of predatory journals which have exploited a heretofore unperceived weakness in the scientific publishing system: the long-standing and multi-faceted ‘gentleman’s agreement’ that constitutes its quality-control backbone.

Science cannot exist without the communication of scientific information. Scientific communication can only be credible if it has an efficient system of quality control. There are no international conventions or laws governing this quality control, which has evolved informally within the scientific community to become the present-day peer-review and editorial processes. The miracle of the scientific enterprise is that it has advanced so well with nothing more than the ‘gentleman’s agreement’ that morally binds authors, editors and reviewers to produce good quality, honest work.

The foundation of the review process is to reduce Type 1 error — accepting a manuscript that is fatally flawed. Type 2 error — rejecting a good manuscript — is considered potentially less harmful to science. While we may quibble about this last point, we must recognize that the only way to ensure that no good manuscript is rejected is to accept all submissions, and this would be fatal to science! Just as in statistics it is impossible to reduce the probability of Type 1 errors without increasing the probability of Type 2 errors, so a rigorous approach to Type 1 publishing errors will invariably been threatened with extinction. The progressive decline in the numbers of ‘gentlemen’, however, together with the intrusion of considerable numbers of the exact opposite in the scientific publication sphere, have combined to now threaten the very existence of science. Predatory OA journals, which by their very nature have no regard for Type 2 errors, have thrown the prevention of Type 1 errors to the winds, and we believe that science, the cornerstone of modern human civilization, now faces an existential threat.

The Prelude

The advent of the Internet has made it possible to de-materialize scientific journals, reducing the costs of publication to the initial outlay of a computer, a few programs, and an annual fee for provision of a website. Online versions of established print journals, as well as the first online-only journals, began to appear in the early 1990s. These were ‘well-intentioned’ journals that adhered to the values and gentlemen’s agreement of the traditional journals.

Given the near-universal access to the Internet, the printing and mailing cost-savings, and the growing expectation of ‘free’ content on the Internet, the ‘Open Access’ publishing model emerged, in which the user (reader) would have free access to all articles, and the administrative costs would be offset by page charges — in effect, authors would pay to have their work published. The embryo of this model already existed in the traditional journals of scientific societies (such as NSA) that provided their journals to members, who typically paid small annual membership fees — ergo the necessity for page charges, to cover the much greater costs of paper publication.

On February 14, 2002, the ease of desktop publishing and the open-access business model intersected with a clearly social and quasi-political message, in the form of the Budapest Open Access Initiative (BOAI - www.budapestopenaccessinitiative.org). 'Open Access' came to embody the 'democratization' of access to knowledge, previously restricted to a perceived wealthy and exclusive scientific elite (www.budapestopenaccessinitiative.org/boai-10-recommendations; www.plos.org/open-access/; openaccess.mpg.de/Berlin-Declaration; blog.scienceopen.com/).

Like most manifestos, the BOAI statement is a cleverly-designed document that substitutes ‘motherhood and apple pie’ dogma for critical thought. It can be summed up in the Vision Statement of the Open Access Academy website: ‘Freely available research results for everyone’ (www.oaacademy.org/vision-and-mission.html) — apparently accustomed to gratuitous luncheons. The BOAI statement exploits human cognitive and moral weaknesses and provides a platform for members of a vocal social movement. Those wishing to announce their adhesion to the open-access social movement simply repeat the ideas and concepts presented in the original Budapest statement or the ensuing copycat statements. The continual and collective repetition of the OA mantras has assumed the status of a consensual truth.

The fact that OA restricted access to publishing scientific information (Frank 2013, Burchardt 2014), and therefore further disadvantaged legitimate scientists with small research funding, both in developed and developing countries, seemed to be lost in the ‘free access’ euphoria. It was proposed that these scientists need only send a letter to the OA journal declaring their impoverished state, and all would be fine. Intentionally or not, it apparently occurred to nobody that this was a demeaning process which would constitute a very real barrier to publishing for many scientists, not least of them from European countries with great pride and small resources.

Setting aside these serious reservations about access, it must be said that the original OA intentions were honorable on the fundamental quality-control issue. The BOAI reads ‘Open access to peer-reviewed literature is the goal.’ In the years following the BOAI, we realized that this was a Panglossian statement, formulated as the OA proponents watched unicorns peacefully graze on the grass outside. While thus distracted, a sinister and previously unknown menace began to appear: predatory journals. These are the scientific journal equivalent to counterfeit coinage; like the latter, they appear to represent a guaranteed value, while containing at best a reduced value, and often virtually none at all. The reason: abandonment of the ‘gentleman’s agreement.’ Predatory journals have no quality control. They have accepted and/or published papers by the Simpson family, by computer-generated nonsense programs, and by sting operations (www.sciencealert.com/twoscientific-journals-have-accepted-a-study-by-maggie-simpson-andedna-krabappel, Bohannon 2013). Their publishers have appeared virtually overnight, like Scientific Online Publishing (SOP), which appeared in 2013, and by 2014 was ‘publishing’ 42 journals, nominally covering a vast spectrum of scientific domains. Public Science Framework (PSF) outdid them, opening instantaneously both of these prerequisites, we instead randomly sampled 65 of the 693 publishers on Beall’s 2016 list of predatory publishers (scholarlyoa.com/2014/01/02/list-ofpredatory-publishers-2014/), computed the mean number of journals per publisher (11), and multiplied this by the number of publishers on the list, to obtain a very conservative estimate of the total number of predatory journals in 2015: 10,153. Applying the procedure above to each year since 2011, the progression of predatory journals is shown in Figure 1. Our measurement technique achieves results quite close to those recently published (Shen and Björk 2015) using a more complex multistage stratified sampling design (7,623 vs. approximately 8,000 predatory journals in 2014, with an estimated 420,000 published articles!). The estimates are conservative because the list includes only publishers of multiple journals, and not ‘stand-alone’ journals, which number 882 to date (scholarlyoa. com/2016/01/05/bealls-list-of-predatory-publishers-2016), and the list is undermanned, and relies upon volunteers to report predatory journals, which are then examined individually.

To date, the Directory of Open Access Journals (DOAJ) lists 11,315 OA journals (doaj.org). Prior to 2016, the only requirement for inclusion in the DOAJ list was that the journal be OA, which obviously qualified many predatory journals. Minimum quality criteria were introduced in 2015, such that the degree of overlap between the journals in the DOAJ database and Beall’s list is currently not known, although if a reader has approximately one year of free time, this could be ascertained. Notwithstanding, the conservative estimate of the number of predatory journals in 2015 was 7,623, or ≈ 75% of the number of DOAJ-listed journals in 2015 (Fig. 1).

The 2014 ISI Journal Citation Reports lists approximately 14,000 journals that are considered genuine (due to re-structuring of the JCR website, we are unable to ascertain the current number). After less than 5 years, there were thus already half as many predatory journals as there were genuine journals, which have built their reputations over the past two centuries (Fig. 1). Although the number of data points is necessarily limited by the novelty of the predatory OA phenomenon (years), and thus far too small to construct a serious model, a very short-term projection suggests that if the current conditions persist, by approximately mid-2016 there will be as many predatory OA journals as genuine journals.

Just as the internet has provided the conditions for predatory journals to flourish, so it amplifies the dilution of science by giving predatory journals a visibility indistinguishable from normal journals, simply because search engines cannot evaluate quality, only recognize words. An analysis of the geographic origins of predatory journal authors, however, reveals that fully 80 percent are from developing countries (Shen and Björk 2015), normally the ones who should have the least money to pay for page charges! Such a situation raises even further questions of credibility and, in a perverted manner, vindicates the utopian belief that OA will increase the developing world’s access to ‘publication.’

In the developed world, scientists can, at best, recognize most of the genuine journals in their fields, but are unable to do so consistently in other fields; understandably, their students at the undergraduate level cannot even do that. While searching for referees for submitted manuscripts somewhat out of our own fields (PGB and SS), we now bring up as many recent articles from predatory as from genuine journals. Student reports now include predatory journals in their reference sections. As the number of predatory publications rises, science is facing a ‘dilution crisis,’ in which science itself will soon be reduced to homeopathic levels.

Read the rest of this article in the June 2016 issue of World Aquaculture Magazine here

Share this:
Tags:

About Peter G. Beninger, Jefffrey Beall, Sandra Shumway

Université de Nantes, University of Colorado, University of Connecticut

Gold Sponsors

JWAS Editorials

  • 2023

  • 2022

  • 2020

  • 2019

  • 2017

  • 2016

  • 2015

  • 2014

  • 2013