Following yesterday's post which touched on the question of blogging referee reports, several commenters on Michael Nielsen and Lance Fortnow's weblogs think this ought not to become widespread because of the chilling effect it could have on reports.

This consideration aside, on a very immediate level, it would seem rather
rude to publish what is normally understood as a private communication
without asking for permission. It would be analogous to publishing a
private email without asking first.


Daniel Lemire thinks reviews should be public from the get-go. There
are indeed journals that use open peer review. The Journal of
Interactive Media in Education, for instance, implements both private and public open peer review processes as follows:




There was an excellent article about open peer review in Nature a
couple years ago, documenting the experiments in open peer review that
were conducted by the journals Atmospheric Chemistry and Physics, Electronic Transactions on Artificial Intelligence, and the Medical Journal of Australia. Recommended reading.


And over at David Wiley's site there's a discussion of whether open peer review indeed improves quality. I suspect it strongly depends on the field, and on which conference/journal you are considering.


Finally, some biomedical researchers have been suggesting to do away with (pre-publication) peer review altogether!
[Seb's Open Research]