I’ve just peer reviewed a paper for the first time. It’s a funny transition, from being at the receiving end of a tough review to writing one myself, and an unexpected change that I’m seeing in my opinions about peer review is this: as an author, I have been agnostic about single-blind vs. double-blind review, but as a reviewer, I would much rather not know the identity of the author.
[Quick note to non-academics: peer review is the process by which scientists decide whether a journal article should be published or not. Peer review is meant to assess the quality and clarity of science being presented in the journal article. Reviewers make recommendations to authors for how the manuscript can be improved and to journal editors for whether the article should be published. In single-blind review, authors do not know the identity of the reviewers, but reviewers know the identity of the authors. In double-blind review, neither author nor reviewer know each other’s identity. In ecology and evolutionary biology, single-blind reviewing prevails.]
In this helpful guide to peer review published by the British Ecological Society, the authors suggest that reviewers watch out for and try to avoid the influence of the following biases:
- gender bias
- geographic bias
- seniority bias
- confirmation bias (i.e. discriminating against results that support controversial or unconventional theories)
Avoiding bias is an incredibly difficult thing to do. In many walks of life, we have no choice but to confront situations in which our biases may arise, and must (or at least ought to) actively consider how to diminish these biases, as difficult as that may be. But peer review need not be one of these situations. The first three of the biases mentioned above can readily be avoided by adopting double-blind peer review. It is much easier to avoid bias when you don’t know who the object of potential bias is. And even if one manages to suppress bias, it is far less stressful to avoid the need to suppress bias altogether.
A major objection to double-blind peer review is that a reviewer can probably figure out the author if they want to, based on the subject of research. Hence the effort involved in double-blinding (which would include not only redacting names and contact information but also modifying citations such as “we have previously shown (X&Y 2007)…”) is not worth it. But the key words here are can and want to; this view ignores the possibility that a reviewer may not bother to find out who the author is, or indeed would actively prefer not to know who the author is. Moreover, as this editorial from Nature Geoscience notes, while it might be easy to identify the lab or research group from which the paper emanates, figuring out the identity of the first author of a paper is not straightforward (this is exactly what would have happened if I had not known the authors of the paper I just reviewed). Being even a little uncertain of the identity of the first author could go a long way in avoiding the unconscious biases that all scientists experience.
Curiously, two of the leading animal behaviour journals, Animal Behaviour and Behavioral Ecology, are among the very few journals in our field to offer double-blind peer review. While the former has flexible policies regarding identity disclosure, it appears that even here, a reviewer cannot choose to remain unaware of the author’s identity if the author chooses to disclose his or her name.
For the sake of completeness, here’s an older editorial from Nature that lays out some arguments against double-blind peer review. I don’t think I understand most of them (e.g. why knowing the identity of the author would help in “differentiating between a muddy technical explanation and poor experimental technique”) . I imagine that my views on this subject might change as I review more often, but right now I don’t see the harm in offering the option to both authors and reviewers for double-blind peer review.