I no longer do anonymous reviews. This page aims to explain why I believe anonymous reviews are wrong and should be eradicated for the sake of ethical behaviour and scientific quality in publications. My point here is to argue that signed reviews are of better quality and make the reviewer responsible; anonymous reviews, on the contrary, allow for dishonest and scientifically irresponsible behaviour.
The reason I started this is that, just like so many others, my work has been subject to several dishonest reviews that it was not possible to counter because the reviewers were protected by their anonymity. Even though I tried to use my right to respond, and did when this was granted, no real (scientific) dialogue was possible. Instead of complaining or believing the world is like this and nobody can change it, I believe the solution is to stop anonymous reviewing. I have stopped doing anonymous reviews and try to convince others to stop as well whenever I am invited to review something or when I sit in program committees.
From a positive point of view, there are various reasons why I believe signed reviews are good: In fact I have spent quite some time writing reviews during my professional life, and have never seen any reason why my identity should be concealed or why I should not interact directly with the authors whose papers I review. Also, in roughly 90% of the cases where the authors themselves were "anonymous" I was able to identify who wrote what, so anonymity is not reciprocal in most of the cases.
The way things are now, doing a good review or a bad (lousy or dishonest) one makes no difference for the reviewer (the only thing which is public is in how many program or journal committtees s/he sits). This is an obvious way to discourage thorough work and to encourage bad practice.
So, I decided that henceforth I will only do signed reviews, which disclose my identity to the authors and which I can make public if I want to (after all, it is my own work) -- note that it is not necessary to show the original text and/or original passages, just comment on them. Even though the second condition may sound like posing several problems (there may be cases where it may be advisable not to do it), just think of the pedagogical advantage: Many things can be learned (and mistakes not repeated) by studying the objections and the problems in previously submitted (accepted or rejected) papers to the conference or journal you are intending to publish in.
Let me illustrate some cases of anonymous reviews I have received myself in order to give a flavour of what can happen. I have of course also received extremely helpful and interesting anonymous reviews (as well as produced some which I hope were taken positively by the authors). I wish I could know who the reviewers were so that I could thank the reviewers personally (and publicly in the paper's acknowledgements).
This is my standard answer, and I encourage as many people as possible to do the same:
Thanks a lot for your invitation.
I do believe, however, that anonymous reviews are not the right thing to do in any case as they
1) fail to give credit for those who write them
2) encourage unethical reviews from dishonest individuals
I therefore only produce signed reviews, for which I take all blame and credit.
If you strongly agree or disagree with the opinions posted above, please send me a note. I will post it here or produce links to it.