The marquee research journal Nature and almost all of its sister publications this week announced that they will offer authors the option of participating in double-blind peer review, where both submitters and referees remain anonymous. The practice, which is common among humanities journals, has long been debated in the sciences, and several journals have recently taken the plunge. Some observers, however, remain skeptical of the value of double-blind systems and note that other journals are heading toward greater transparency.
Traditionally, scientific journals have adhered to a single-blind system, in which authors don’t know the identity of those reviewing their paper. But that system has led to concerns that it may contribute to bias against some authors, including women, minorities, and those from less prestigious institutions. In the last decade, publishers have tried to address those concerns by introducing various tweaks to the reviewing process.
Nature Publishing Group (NPG) began testing the double-blind system with Nature Climate Change and Nature Geoscience in May 2013, after an author survey indicated significant interest in the model. “We’re here to serve the needs of the research community, and it’s become increasingly clear that they want to have the option … to choose double-blind peer review,” said Véronique Kiermer, director of author and reviewer services at NPG.
“The trial in Nature Climate Change and Nature Geoscience gave us plenty to be confident about,” she says, but notes the option of complete anonymity will be “an ongoing experiment.” NPG “may find that different disciplines react in different ways,” she says, “and as practices evolve in the future, we will want to evolve with them.”
Editors at other journals, including Science (publisher of ScienceInsider), are watching the experiment at Nature and elsewhere, but haven’t made the jump. In part, that’s because some have concluded that it would be hard to prevent reviewers from correctly guessing who authors are, particularly in small fields. Some reviewers can predict authorship by looking closely at a paper’s references; authors often build on their previous work and thus cite themselves extensively.
“With double-blind, the inevitable guessing game will begin of reviewers trying to guess which group authored the research,” wrote Marcia McNutt, editor-in-chief of Science, in an e-mail. One upside to switching to a double-blind system, she notes, could be a lower number of self-citations (which are often frowned upon by journal editors). And she adds that “it will certainly be important to determine if double-blind improves … equality for women authors in the process.”
“There is pretty good evidence for various biases in the way that articles are perceived by reviewers … and double-blind review is one possible way to avoid, or at least mitigate, these biases,” says Michael Eisen, a biologist at the University of California, Berkeley, and co-founder of the nonprofit open-access publisher Public Library of Science (PLOS). But, so far, PLOS has concluded that it is too difficult to mask authors’ identities in fields such as biomedicine, he says.
César Hidalgo of the Massachusetts Institute of Technology in Cambridge is skeptical of anonymity in peer review. “Unblinded reviews,” he says, are better “because they equalize the power between the author and the reviewer.”
To address such concerns, some publishers are moving to highly transparent reviewing systems, in which both authors and reviewers are identified. And a few journals, such as F1000 Research, go even further by publishing referee comments alongside a paper and making the comments searchable and citable. The idea is to give referees, who generally work for free, some public credit for their efforts. But critics of open peer review fear it can also cause image-conscious reviewers to be less critical.
Publicly recognizing reviewers won’t be possible in Nature’s system, but Kiermer notes that, “at the moment, reviewers can obtain a certificate of their reviewing activity for Nature journals.” And the publisher is continuing “to consider some form of open review as an option for the future, in response to author and reviewer feedback,” Kiermer says.