Maintaining high research integrity standards at PLOS ONE
Since 2006, PLOS ONE has published >200,000 articles, providing an inclusive home for primary research spanning all scientific disciplines and representing researchers from around the globe. As reflected in the journal’s publication criteria and policies, PLOS ONE is strongly committed to upholding high ethical standards for the conduct and reporting of scientific research. We are a member of the Committee on Publication Ethics (COPE), and we welcome readers, authors, editors, and reviewers to write to the journal office if they have concerns about the scientific validity or integrity of our submissions or publications. Journal staff follow up on all concerns raised to the journal per COPE guidelines.
The journal has always worked to implement systematic approaches to identify and address publication ethics concerns in our submissions. For example, we use software tools to highlight text overlap with published works, similarities to previous PLOS ONE submissions, and authorship changes during peer review, and we require authors to make underlying data available to enable full transparency as to the basis of published results. In recent years, we have centralized and increased editorial resources devoted to research and publication integrity, and in January 2018 formed a dedicated Publication Ethics Editorial team. This team currently includes three Senior Editors who have scientific research backgrounds as well as specialized expertise in research integrity standards, policies, and workflows. Our Publication Ethics Editors address ethical and scientific concerns involving PLOS ONE submissions and published articles, and also aim to raise awareness of publication ethics standards, develop policies and processes by which we can address integrity issues (at scale) prior to publication, and participate in broader discussions with researchers and publishers pertaining to publication ethics policies and practices.
Figure 1. Distribution of active publication ethics cases at PLOS ONE. Note that cases are categorized by potential issues raised, not by outcomes and/or whether the concerns were validated during follow-up. These data include pre- and post-publication concerns. Dual pub/sub, dual publication or dual submission concerns.
As the journal has developed and grown, we have seen a progressive increase in the volume and breadth of publication integrity concerns raised to our attention. This may reflect increases in the community’s awareness of publication and research ethics issues, researchers’ utilization of software tools to evaluate the integrity of published work, and/or trends toward more open discussion of integrity issues on social media and other online forums. Concerns have been raised for only a very small fraction of PLOS ONE articles, but even so, given the scale of PLOS ONE’s publishing output our caseload provides an informative snapshot as to issues that currently impact scientific research reporting. We receive queries across a broad range of ethical and scientific issues (Figure 1), with the most prevalent – comprising roughly 40% of our active cases – involving image and data concerns (e.g. potential or suspected image manipulation, duplication, data fabrication). In many cases raised to the journal, corrections, comments, or additional discussion can address the issues raised and serve to correct or clarify the scientific record. That said, in cases where concerns are not satisfactorily resolved we take action on the article as per COPE guidelines so that future research will not be built upon a shaky foundation. Over the past two years, 38% of publication ethics and other post-publication cases that we resolved did not require any formal editorial action on the article in question; 41% of pre-publication cases resulted in rejection of the submitted manuscript; and 64% of post-publication cases resulted in retraction, correction, or an expression of concern (Figure 2).
Figure 2: Summary of pre-publication and post-publication case resolutions over a two year period. “Resolved” indicates that the issue was resolved without need for a formal editorial action on the article in question such as rejection or a post-publication notice (comment, correction, retraction, expression of concern). EOC, expression of concern.
These data highlight the importance of efforts by the research community and publishers in identifying and addressing concerns in the scientific corpus, including after an article’s publication. Whereas some issues ought to have been identified during peer review, in many cases the concerns raised to the journal may not be covered by the peer review process (e.g. data ownership and authorship issues), or could easily be missed by the editor and reviewers. Whereas 3-4 researchers may evaluate an article during pre-publication peer review, the same article may be read by hundreds or thousands of researchers after publication. As the article becomes organically subject to a substantially greater level of scrutiny post-publication, it becomes increasingly likely that any issues will come to light. This continued evaluation of published research is an integral part of the scientific process whereby we learn from, build on, and improve upon published research, and it is not surprising that it occasionally reveals errors or other issues that need to be rectified. Indeed, sometimes authors themselves notify the journal of issues they realize only after publishing their work which may warrant correction or even retraction. Amendments to the published record to address ethical or scientific issues – regardless of when or by whom they are raised – serve to provide the community at large with a reliable body of literature on which to draw for future research.
Reassuringly, only a minority of issues raised to our attention involve more serious potential research misconduct (e.g. data fabrication). More often, we find that the problems arise due to issues involving researchers’ understanding of best practices in research reporting and the journal’s editorial policies. Based on our experiences, we believe that further guidance in these areas may help to address some of the issues that we encounter most frequently in research integrity cases. The PLOS ONE Publication Ethics Editors are working to develop resources for authors, reviewers, and editors to increase awareness of specific issues and related journal policies, for example in image preparation. By working together with researchers and helping to raise awareness, we aim to help authors strengthen the presentation and reporting of their work, preclude future post-publication issues, and empower Academic Editors and reviewers to identify issues during the peer review process. In parallel to these efforts, PLOS ONE is working together with the research community to develop and update our policies regarding research reporting standards. We value the research community’s insights about these policy issues, and invite you to email us at email@example.com with suggestions as to resources or issues that you feel we ought consider as we develop these policies.
In addressing research and publication ethics concerns, our primary objective is to ensure the integrity and validity of the research we publish. This work can touch on some delicate issues, and we are well aware that the outcomes of this work can have direct repercussions for the researchers involved, particularly if we need to consider retracting an article or requesting an institutional investigation. We take care to provide due process and careful consideration in each and every case, although this can take some time. These are complex issues that require input from multiple contributors – including journal staff, authors, Academic Editors, external reviewers, and institutional representatives – who spend considerable time and effort addressing our queries, providing their expert perspectives, re-assessing articles, providing and/or poring over underlying data files, and investigating concerns at the institutional level when needed. We sincerely thank all who notify us of concerns about PLOS ONE articles and submissions, as well as the many contributors who work with us to address, assess, and resolve these issues. Together, through these efforts, we continue to maintain high ethical and scientific standards at PLOS ONE.
Featured Image: Magnifying glass Julo Wikimedia Commons Public Domain