But increasingly journals are finding out that those supposedly authoritative checks are being rigged.
In the latest episode of the fake peer review phenomenon, one of the world’s largest academic publishers, Springer, has retracted 64 articles from 10 of its journals after discovering that their reviews were linked to fake e-mail addresses. The announcement comes nine months after 43 studies were retracted by BioMed Central (one of Springer’s imprints) for the same reason.
The blog Retraction Watch, which monitors and reports on retractions for fraud, plagiarism and other dubious practices in the academic publishing industry, says this latest announcement brings the tally of papers withdrawn for faked reviews up to 230 in the past three years. Those papers make up only a fraction of the hundreds of thousands of studies published each year, but they have the publishing world concerned.
Retraction Watch co-founder Ivan Oransky, a science journalist with a medical degree from the New York University School of Medicine, said he didn’t know of any instances of retractions for faked peer reviews before 2012. Since then, the practice has accounted for 15 percent of all retractions logged by his site.
“It’s like a virus that maybe was lying dormant for decades or centuries and all of a sudden, it’s coming out,” he told The Washington Post. “What’s not clear is, are we better at finding it? Or is it actually a new phenomenon?”
It’s also unclear who exactly is responsible for the rigged reviews. In a report for the journal Nature last fall, Oransky and his colleagues told the story of a Korean medicinal plant researcher who wrote peer reviews for 28 of his own papers. In July, the publishing company Hindawi found that three of its own editors had subverted the process by creating fake peer reviewer accounts and then using the accounts to recommend articles for publication. (All 32 of the affected articles are being re-reviewed.)
But investigations into suspicious-sounding reviews have also uncovered a number of services selling names and contact information for made-up experts guaranteed to give an expedited, positive review.
In a statement on its Web site in February, the Committee on Publication Ethics (COPE) detailed these agencies’ “systematic, inappropriate attempts” to manipulate the process.
“Some agencies are selling services, ranging from authorship of pre-written manuscripts to providing fabricated contact details for peer reviewers during the submission process and then supplying reviews from these fabricated addresses,” it said. “Some of these peer reviewer accounts have the names of seemingly real researchers but with email addresses that differ from those from their institutions or associated with their previous publications, others appear to be completely fictitious.”
Many academic journals contract with third-party reviewing services that help track down experts to read submissions, and it’s not too hard for unscrupulous reviewers to exploit loopholes in the system. In 2014, SAGE Publishers retracted 60 articles from the Journal of Vibration and Control after uncovering a “peer review ring” in which researchers reviewed their own and each other’s papers under aliases.
The translation, “language polishing” and editing agencies that some authors hire to help get their research published may also be in on the game: The COPE statement said it’s not always clear whether authors were aware that these agencies were rigging their reviews.
“The question here is who knew what, when?” Oransky said. “You have to be thinking about the steroid defense … like the baseball players who say, ‘Well, I don’t know. The trainer told me this was a totally legitimate enhancer,’ but actually the trainer was putting something illegal in it. The question is, did the player actually know?”
The announcement from Springer about the 64 retracted articles is brief — it does not say which papers were withdrawn or where they were published. Nor does it identify the source of the fake reviews, though in a statement to Retraction Watch, Springer said that third party agencies may have been involved.
Publishers are starting to implement policies aimed at preventing fake reviewers from accessing their systems. Some have stopped allowing authors to suggest scholars for their peer reviews — a surprisingly common practice. Many are mandating that peer reviewers communicate through an institutional e-mail, rather than a Gmail or Yahoo account. And editors at most journals are now required to independently verify that the peer reviewers to whom they are talking are real people, not a fabricated or stolen identity assigned to a fake e-mail account.
But some publishers and ethics organizations, including COPE, have suggested that controlling the apparent rise in peer review manipulation requires more than stricter publishing practices. They link the problem to the oft-lamented “pressure to publish” in the world of research.
“There are some very fundamental problems within the reward system of academia which even the best practices in publication ethics cannot solve unless the fundamental issues in academia are addressed,” COPE’s chair Ginny Barbour wrote in December. “The uncovering of companies systematically manipulating publications, by the use of fake reviewers and more, offers an alarming glimpse into what can happen if reward systems are implemented with no thought or oversight.”
And Oransky said that when he examined some of the papers retracted for rigged reviews, they didn’t seem to have any other problems — most of them likely would have been approved by a competent reviewer he said, though perhaps not as effusively or as fast.
“This almost seems like more of an insurance policy than making the difference between rejection and getting accepted,” he said. “That speaks to the pressures that researchers are under.”
He continued, “Publishing papers is the coin of the realm when it comes to academic advancement. … You’re going to do whatever you have to do to do it.”