Blind faith

Journal name:
Nature Climate Change
Volume:
3,
Page:
525
Year published:
DOI:
doi:10.1038/nclimate1923
Published online

Many believe that double-blind peer review reduces referee bias, real or perceived.

Beginning this month, this journal (along with our sister journal Nature Geoscience) is going to trial — as an option to authors — double-blind peer review. Under this system, reviewers will keep their anonymity, as before, but they will not be told whose work it is that they are being asked to review. This contrasts with the single-blind peer review process currently used by the Nature family of journals, including Nature Climate Change, under which reviewers know whose work it is that they are looking at.

Anonymity allows referees to comment candidly on research without undue fear of social or professional reprisal. However, if they know the identity of authors, there is at least the possibility of reviewer bias. This could be for reasons of perceived reputation, competitiveness, seniority, nationality, gender, ethnicity or a whole host of other factors that can easily be gleaned from a manuscript's author list. Most referees — we trust and believe — try to be objective in their assessments, and are not swayed by factors associated with authorship. But simply withholding author information should largely remove the possibility of such bias, whether conscious or not. In so doing, this double-blind process has the potential to make scientific peer review, well, more scientific — or so many critics of conventional single-blind review assert.

We are aware, of course, that double-blind peer review also has some possible shortcomings. Some candidate referees may be reluctant to review manuscripts of unknown provenance; others may in any case know, or correctly guess the authors' identities from the nature of the research under review. Nor can we be sure that double-blind review will make any practical difference in terms of final editorial decisions.

It is also important to stress that the double-blind peer review trial will not be run as a controlled experiment. Participation will not be mandatory, which necessarily means that participating authors will be a self-selected group. We hope, however, that uptake will be good and that the trial will indicate the possible benefits of double-blind review.

There will in due course be a new section of our guide to authors to explain the process, but it will actually be very straightforward. When submitting a manuscript through our website, authors will need to select the option 'Yes I do want to participate in double-blind peer review'. This will prevent authorship and contact information being transmitted to referees. Obviously, authors will also need to ensure that none of the uploaded files contain information that would inadvertently reveal their identity. Details of what to exclude will also be listed in the guide. We hope you will appreciate having this option, and we will be monitoring uptake and soliciting your feedback over coming months.

Additional data