There are two different options of reviewing a manuscript: (1) In single blinding, the reviewer knows the author, but the author does not know the reviewer. (2) In double blinding, both are anonymously for each other. Most of the jorunals use single blinding, but there are also journals available using double blinding.

I strongly advocate for a double blind peer review process to help reduce bias. I can't think of any reason why a reviewer should know the author or institution. Please refer to the following text: "Country development and manuscript selection bias: a review of published studies".

This is exactly the point for the discussion. There is a definite bias in reviewing a manuscript by a established well connected group vs a new investigator from an unknown place ! It should be double blinded for an honest unbiased review

Dear all,

I agree that the problem of bias against new or unknown investigators is potentially serious and that this is a strong argument for double-blind peer review. However, in my view there are some further factors to consider.

(1) First of all, here (http://www.jstor.org/stable/2006906, Blank 1991) is some evidence that bias may be weaker than anticipated, thought there is in fact signficant bias against authors from nonacademic institutions. Also, this study found that "referees are more critical when the reviewer is unaware o f the author'sidentity" - so this would be a further argument against single-blinding. However, another study (Goodlee et al. 1998, http://jama.jamanetwork.com/article.aspx?articleid=187748) found the opposite, i.e. that rviewers who were blinded to authors' identities were less likely to recommend rejection than those who were aware of the authors' identities". So probably before we can dedice, more studies (newer and from other disciplines) are needed to see whether there is any consistent pattern.

(2) The bias is not necessarily only caused by the reviewer but may also - perhaps to an even stronger degree - be on part of the editor. This effect is probably stronger in journals with high rejection rates, where articles are often not even sent out for review, but I am not aware of any evidence for this editorial bias against particular groups of scientists - nor of any practical way to introduce double-blinding in the communications between authors and editors.

(3) When the authors are known to the reviewer, they can more easily check whether the authors self-plagiarize or whether there is any novelty in comparison to what the group has published before.

(4) In many cases, the idenity of the authors may not be concealed effectively, because the system they are working on is already published, or e.g. because the location of their field experiments needs to be named and reviewers can deduce who the authors are. In that case you would get a different bias: well connected reviewers knowing their partners and competitors well more easily guess who the authors are. This is probably not desirable either.

(5) Some journals (e.g. Annals of Applied Biology) do have forms in which the reviewers have to declare any conflicts of interest. This is a compromise to reduce bias but of course is based on honesty. In any case, I think the editors have the responsibility to select reviewers who are unlikely to be biased.

(6) Many, if not most studies that are rejected from one journal will end up being published somewhere else. But the recognition of the paper by peers will then not only influenced by its quality, but also by the affiliation of the authors or the journal where it is published. I.e. an even stronger bias than in the peer-review and prior to publication might be located in post-publication. This is not necessarily an argument against double-blinding (on the contrary), but it also shows that there are more problems elsewhere.

(7) There is always the possibility to challenge the reviews. If the review is of bad quality you can contact the editor and argue against specific points raised by the reviewer. I.e. if the recommendation to reject an article would purely be based on knowing the author or its affiliation, it would be relatively easy to challenge the (weak) arguments brought forward by the reviewer. Though in that case (as I have experienced) it would be unlikely that the editor changes his or her opinion. But again, it is then the responsibility of the editor, whatever the review: single or double blind.

@Thomas. Well detailed and comprehensive answer. Thank you. Strongly agree on point 2. which is more relevant in high rated journals who do not have a democratic way of selecting papers. Its solely at the discretion of the editors.

And, what about a third option: a reverse single-blind review. The authors' names and affiliations are anonymous but the reviewer is not.

I just submitted a lengthy, critical review with a journal and felt, in fact, honored to witness the evolution of the authors' study through the peer review process. I waived my rights to anonymity because 1) the authors didn't have such liberty, and 2) I felt more accountable for providing quality feedback.

Waiving anonymity of reviewers (reverse single-blind) might incentivize constructive and high quality manuscript reviews, where the double blind may not. It may also serve to mitigate bias against the submitting authors.

Also see:

http://www.uvm.edu/~pbierman/classes/critwrite/2013/peer_reviewing.pdf

Article: Anonymous Reviews: Self-serving, counter-productive, and unacceptable.

I agree with #4 in Thomas' post above. Double-blinding turns out to be somewhat of a joke for many of our submissions because we're one of 5 groups that have published work with a particular technique, so we get the bias of the reviewers without the benefit of knowing the reviewers. I just published with Frontiers and found the experience of not knowing who the reviewers were in the interactive phase of the review but finding out their identities at publication to work pretty well, but I don't know what happens if the manuscript is ultimately rejected, whether in that case you find out who the reviewers were.

I think the anonymity protects the more early-career reviewers so reviews appropriately leading to rejections don't lead to retributions, but can also lead to abuses and biases by those currently favored by the system. A leading journal in my field recently published a report that as far as publishing in it is concerned, "Success breeds success," meaning a large proportion of its published authors are published in it repeatedly. I'm not sure how to interpret this report. Does success actually breed success or does it require name recognition/being well-connected to get published in this journal? Probably a mixture of both, but I do wonder sometimes what proportion of acceptances are attributed to the latter factor, especially when I see published in this journal many articles of similar or lesser merit than the ones we submitted to it and received rejections for (while simultaneously being requested by said journal to review submissions "because of your expertise" which the journal presumably knew about from my previously rejected manuscripts). . . the irony!

I don't know about other fields' practices but I haven't seen a single journal within the field of information systems that doesn't use double blinding. The same applies for our conferences as well. I have never reacted negatively to it.