29 July 2010

Nature didn't want it, so you get it

The following is a letter I submitted to Nature in response to a recent editorial. They decided not to publish it... so... blog fodder it is!


I was pleased to read the recommendation that “scientists, institutions and funding agencies must increase transparency wherever possible” (Nature 465, 7; 2010). To that end, I suggest journals such as Nature consider ways to increase the transparency of the peer review process. The practice of anonymous peer review is at odds with the increasing adoption of transparency and accountability by government, private industry, and elsewhere.

Currently, the identity of reviewers typically remains anonymous to the author(s) of the paper during the editorial process. Even after the peer review process is over and the final article of record has been published, reviewers remain anonymous to the readers. This level of anonymity makes it easy for authors or readers to claim that a shadowy cabal of insiders can effectively block publication of science that is technically sound but controversial. Such accusations were raised by some stem cell researchers earlier this year (http://blogs.nature.com/news/thegreatbeyond/2010/02/bad_reviewers_block_good_resea.html). Cases like it show how a lack of accountability in peer review can be used to cast doubt upon scientific evidence, particularly for controversial subjects like climate change.


Unknown said...

what are your thoughts on the EMBO Journal practice of publishing the reviews (while keeping them anonymized). For what its worth, I wonder whether non-anonymized review would be better.

Zen Faulkes said...

It should help, methinks.

That said, I haven't read EMBO Journal lately, so I haven't seen how they do it in practice. I could see some oddities cropping up. Reviews are often made on only the first submission of a paper, and many of the reviewers' comments may have been addressed in the final archival version. So it strikes me that it could be a bit confusing.

Also, some comments do take on a different spin if you know who's making them. It sounds different if I say, "You should cite Faulkes, 2010," compared to when someone else says it. (Me being the aforementioned Faulkes and all.)