just out of curiosity, what could be the reasonable expected purposes for requesting a single-blind review instead of a standard double-blind in your model?
best,
dj
On Tue, Nov 6, 2012 at 3:56 AM, Chitu Okoli Chitu.Okoli@concordia.cawrote:
Actually, I think it is more reasonable to use double-blind by default unless authors request single-blind. If single-blind were the default, it would be difficult to request double-blind as exceptions:
- If there is a "big name" researcher who wants to take advantage of
his/her reputation, he/she would not request double-blind.
- If there is a "big name" researcher who is modest and does not think
highly of himself/herself, he/she would not request double-blind.
- If there is a minority or woman researcher afraid of discrimination, if
he/she requested double-blind, the reviewers would reasonably guess that the author(s) are minorities or women.
Thus, I think double-blind as a default for everyone with single-blind as special exception would be the more practical and fairer general policy. With the increase of preprints and working papers (e.g. arXiv and SSRN), I think author anonymity is becoming increasingly impractical.
In any case, these comments mainly apply to social science journals; I still think that single-blind makes more sense for computer science journals.
~ Chitu
Kerry Raymond a écrit :
I think a compromise position is to use single-blind unless the authors request double-blind and are therefore prepared to undertake the "ridiculous gymnastics" required.
Certainly (in computer science) I would agree that it is very hard for any established researcher publishing in their normal field to successfully disguise their identify.
Sent from my iPad
On 05/11/2012, at 8:30 AM, "Chitu Okoli" Chitu.Okoli@concordia.ca Chitu.Okoli@concordia.ca wrote:
Although in theory double-blind review is superior to single-blind, in practice double-blind vs. single-blind review has very little to do with journal quality. It is much more a matter of disciplinary culture. (Single-blind: authors don't know who the reviewers are, but reviewers do know who authors are; Double-blind: neither authors nor reviewers know who each other are)
Yes, I am certainly aware of studies that show that double-blind reviewing does indeed reduce the bias towards publishing famous researchers and reduces the bias against publishing work by minority researchers and women. So, I believe that double-blind reviewing is indeed meaningful. However, my general observation is that the decision for a journal to be double-blind or single-blind is mainly a matter of disciplinary tradition. To make a very gross generalization, social science journals are generally double-blind, whereas natural science and mathematical science journals are generally single-blind. This observation is very relevant to this discussion, because the wiki-research-l community sits in between this divide. My perception is that 90% of people who post on this list are in information systems (double-blind), computer science (single-blind) or information science (either double- or single-blind).
If we can accept that single-blind journals can be considered as high-quality as well, then I feel a journal with wikified peer review would do much better being single-blind, especially if its subject matter is wiki-related topics. I understand that one of the primary reasons many journals decide against going double-blind is because of the ridiculous gymnastics that have to be undertaken in many cases to try to hide authors' identity. In computer science, where many researchers make their programs available on the Web, and much of their research concerns particular websites that they have developed, double-blinding would often be impossible if attempted--reviewers couldn't properly evaluate the work without knowing who created it. I think this is very much the case for a wiki-based peer review system for much wiki-related research.
Of course, the official reviewers should remain anonymous. (I know that open peer review has often been proposed--authors and reviewers know each others' identities--and has even been experimented with on several occasions, but it does not seem to be a quality improvement--I can post citations if anyone is interested.) It is much easier to hide the identity of the official reviewers than it is to hide that of the authors, and the benefits of single-blinding are substantial and proven.
~ Chitu
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l