What’s Not Being Discussed, or Considered, in Science Publishing?

To the Editor,

The December, 2014, special issue of the Journal of Microbiology & Biology Education (JMBE) is an extremely important and pertinent collection of ideas related to responsible conduct of research (RCR), ethics, and research integrity. It comprises a wide range of topics and viewpoints that focus on core issues affecting every scientist, many of whom also perform the function of educators. Science and science publishing are in great turmoil, and this special issue is thus extremely timely.

In this brief letter, I wish to focus on a few aspects that I believe were not fully covered by that special issue, in a bid to expand the ongoing discussion, to tie in more factors that can perhaps influence these three core issues that were covered by that special issue, and to ultimately raise more awareness. The greater the awareness, the more open the discussion. The more open the discussion, the faster we can advance the action that is needed to evolve in order to address the pressing issues affecting all scientists and educators. And the faster we can advance these issues, the closer we might be to achieving the more ideal “integrity mindset” that C. K. Gunsalus (p. 120–123) alludes to. I should add, by way of a disclaimer, that I do not consider myself in any way to be an expert in any of these fields, and that my ideas and perceptions exist simply from the eyes of a scientist at the grass-roots level who is involved with publishing, editors, journals, and publishers daily, and intensively exposed to the issues and the conflicts that are associated with these entities.

There appears to be—at least in my perception—an excessive focus on the responsibility of authors and not enough focus on editors or publishers, who also have a responsibility within the publishing process in maintaining the integrity of the published literature. In this letter, I wish to focus on a few aspects that might constitute a gap that may be skewing reality. All scientists who submit papers to journals must abide by the publisher’s rules and ethics, closely following the clauses indicated in the instructions for authors. Yet rarely does the authorship base see similar clauses being applied to the editors, and even less so to the publishers. While the authorship base should indeed be carefully scrutinized, and held accountable for what is submitted, so too should editors (and publishers) be held accountable for what they have allowed to be published in their journals. What appears to be an increase in retractions—or at least a greater awareness of them—as indicated by A. Marcus and I. Oransky (p. 151–154), and their economic impact on the economy (D. B. Resnik, p. 159–161), indicates that what was once perceived to be a robust and fail-safe peer review system is now clearly starting to show cracks in its infrastructure. In some cases, errors in the literature may have arisen because of misconduct, but in other cases, a permeable, porous, and imperfect peer review (9) is what allowed errors to be ultimately approved (by editors), and published (by publishers), despite their genuine desire—in most cases—to ensure an error-free literature. A retrospective—and, more rarely, introspective—review of the literature, most commonly through post-publication peer review (PPPR), tends to reveal those gaps, leading to errata, corrigenda, and, in worst-case scenarios, retractions (6). As sites such as PubMed Commons (http://www.ncbi.nlm.nih.gov/pubmedcommons/), PubPeer (https://pubpeer.com/), and Retraction Watch (http://retractionwatch.com/) gain traction—not without their fair share of controversy—so too, the focus on editors and a potentially imperfect system that science, technology, and medicine (STM) publishers have in place begins to increase. It is here that E. Wager (p. 146–150) aptly calls for a more robust system by editors to deal with “evidence of possible misconduct from a number of sources including peer reviewers, readers, and whistleblowers.” However, misconduct is the most extreme of the issues that an editor would have to deal with (11), and thus the call for a robust system by Wager is valid. Nevertheless, how does that system take into account more subtle problems, such as dubious data, questionable gels, or simple oddities or irregularities about scientific aspects of a paper? The Committee on Publishing Ethics (COPE) code of conduct for journal editors currently in place indicates that editors have the responsibility of following up on claims made about journal articles, even those that are made anonymously (1). It is in this respect, perhaps, that greater accountability and responsibility by editors, and publishers, is required. It is a complex issue, no doubt, especially since the peer pool is already extremely strained, and most editors complete their task voluntarily, so having to deal with additional issues over and above regular editing may overburden the currently strained system in place by most STM publishers. Yet, should editors and authors who are made aware of errors in the literature, even if anonymously, but choose to not correct that literature not be held more accountable? It is likely for this reason that, in a bid to make editors more accountable for upholding the scholarly content of their journal, there is an increasing movement for editors to sign a code of conduct: “Ethical Practices of Journal Editors: Voluntary Code of Conduct” (https://editorethics.uncc.edu/editor-ethics-2-0-code/; Dec. 2014). By holding editors fully accountable, including for past errors in the already published literature, and by simultaneously creating a system that expects greater responsibility by or from editors—a system that will fail if there is no follow-up and effective implementation—it may be possible, through PPPR, to begin to systematically correct the literature that has documented problems (e.g., cases at PubPeer or PubMed Commons). The issue of anonymity is complex because it carries a negative connotation associated with whistle-blowing (12) but may also carry untold benefits, as demonstrated by hundreds of cases already documented at PubPeer and Retraction Watch. Some of these have led to investigations by journals and research institutes, resulting in corrections and even retractions (7). So, the issue of anonymity needs to be discussed, although excessive or intrusive regulation might hypothetically stifle the natural flow of PPPR. Related to this aspect, even as this letter to JMBE was being processed, a new, and potentially controversial document was released by COPE on March 17, 2015 (3).

The impact factor (IF), as well as new emerging metrics to quantify author, journal, and publisher productivity and performance, or ranking, are non-academic factors. Not because their principles are wrong, or because their objectives are not noble, but simply because such systems are open to being gamed, and thus abused, hence potentially corrupting academia. Authors who might be rewarded proportionally based on their IF scores, as occurs officially in many countries around the world, might seek to find ways to obtain a paper in a higher IF journal; similarly, journals and publishers, aware of the “game,” might seek ways to cut editorial corners (2), and speed up publication at the expense of quality control, all in a bid to increase their IF scores and hence the flow of authors to their journals. Thus, a science publishing system that uses the IF as its central axis to evaluate quality and performance, and subsequently reward scientists, is fundamentally flawed. This key aspect was not apparently considered by E. C. Mah (p. 143–145) when calling for greater interaction between journals and institutions to “promote responsible conduct.” If the “system” is built around players whose objectives are being influenced by non-academic factors like the IF, even an endless amount of interaction between these players will never resolve the issue, simply because the potentially wrong incentive is in place. It is precisely for these reasons that DORA (http://am.ascb.org/dora/), the San Francisco Declaration on Research Assessment, continues to call for the removal of the IF to assess the productivity of science and scientists.

On the issue of authorship and authorship guidelines, the Vasconcelos et al. paper (p. 155–158) unfortunately does not discuss the four-clause ICMJE guidelines pertaining to authorship (4), one of the most central documents in science publishing today. These guidelines are neither trivial nor redundant, simply because the main STM publishers officially endorse—and implement—these guidelines, which serve as a layer of “ethics” in science publishing. Despite this, some STM publishers continue to employ an outdated three-clause ICMJE definition. The fourth clause, which relates to public accountability by authors as to the “accuracy or integrity” of their work, is absolutely fundamental in ensuring RCR.

Scientists can no longer simply focus on their legacies. Issues in science publishing, errors in the literature that are waiting to be corrected, and concerns about aspects of STM publishing, including gaps in the responsibilities of editors and publishers, should serve as a call for scientists to take a more pro-active role in contributing to improving a system that clearly has—as evidenced by the wide and healthy discussion in the special issue of JMBE—much to be improved. It is only when more topical issues, as well as less palatable ones, are considered in unison that we may begin to inch closer to achieving integrity in science publishing (5) and RCR, to finally claim that the system is sound and ethical.

Finally, two hot topics requiring urgent attention are not being sufficiently discussed among the mainstream publishers or debated among scholars. The first is the need to publish negative results as a complement to the positive ones, a process that will require a change in the mind-set of the publishing establishment (8). The second is the associated risk of having supplementary data, including negative results (or not-so-positive ones) available as open access online files (10).

I am thankful to Samantha Elliott, the Editor-in-Chief of JMBE, for providing me with this unique opportunity of sharing my views.

Sincerely,