In recent months, several academic publishers have been forced to retract a combined total of over 250 published articles across a number of disciplines due to irregularities and outright fraudulence in the peer-review process. As a result of this, the stringency of checks carried out by publishers during the submission and review process of manuscripts has come under the spotlight. Here, we give some background to this ongoing issue and discuss the various practices by which it can be contained.
At the beginning of 2014, several journals under the aegis of a major publishing house were forced to retract 100+ papers when it was revealed that these papers had been “created” by SCIgen, a computer program that generates ‘random Computer Science research’ using context-free grammar. Later that year, a piece in Nature by retractionwatch.com activists Ivan Oransky, Cat Ferguson and Adam Marcus warned of the increasing danger of authors surreptitiously reviewing their own papers; an issue taken up by COPE (Committee on Publication Ethics) in astatement on 19 December 2014. Although ostensibly different issues, they both pointed towards problems residing at the core of academic publication – the soundness of peer-review.
In March of this year, a Web of Science-indexed cell biology journal was put “on hold” by its publisher following revelations of a “pattern of inappropriate and compromised peer-review”. Later that month, another major publisher issued a statement concerning the retraction of a number of papers (forty-three, according to Retraction Watch) over fears of fraudulent peer-review; possibly involving third-party companies selling their services as false reviewers. The latter publisher, BioMed Central (BMC), having identified a number of suspect email accounts, subsequently offered to share this information with others in the interest of openness and to help stem the problem. This led in the following months to retractions from a number of other journal publishers, as detailed by Retraction Watch here, here and here. As that watchdog noted, the number of retractions due to falsified peer-review was “officially becoming a trend.”
Frontiers was one of the recipients of the list of email addresses of false reviewers provided by BMC. Following a thorough cross-check of these names and accounts with our author and editorial board database, it was confirmed that not one of the false reviewers had reviewed for our journals. And, while this issue remains critical, Frontiers have a number of checks and balances to ensure the soundness of our peer-review and to rule out – as much as is possible – attempts to “cheat” the system through fraudulent peer-review; as well as simply ensuring that ethical issues and potential conflicts of interest are identified and resolved at an early stage.
In addition to utilising iThenticate to screen submitted manuscripts for possible plagiarism, checks are also in place at submission stage to identify potential conflicts of interest between authors, reviewers and handling editors (e.g. familial or affiliations links, recent co-authorships). Frontiers also plans to integrate ORCID into its Loop platform by the end of 2015. Most importantly, the identity of reviewers who endorse publication, as well as that of the handling editor, is disclosed on every paper published at Frontiers for maximum transparency and accountability. In order to partake in the Frontiers review system, a researcher must be registered therein. As a member of COPE, we also refer to that organisation’s guidelines, where necessary, when potentially ethical issues arise during manuscript review. While Frontiers has a number of systems in place to identify potential fraud in the peer-review process, we are certainly taking recent developments among other publishers very seriously. And, while internal investigation has not found any cases similar to those experienced by others noted above, going forward we are working with a number of publishers and organisations to share experiences and data in order to tackle this issue.