On Thu, 16 Mar 2006, C.Oppenheim wrote:
> Andy is correct. RAE panels don't give a toss whether the paper has been
> peer reviewed or whether it has appeared in a traditional refereed journal
> or not. Thus, Stevan's comment that " the RAE ....[do not] regard.....
> unpublished papers as published papers." is incorrect, as the panels take no
> view on this matter at all. RAE panels evaluate the output on its merits as
> a research document. But it would be a very high risk strategy to submit
> non-refereed material as part of one's RAE submission.
This is too tempting not to lure me out of my shy retirement:
(1) The RAE panels don't give a toss whether a paper was published in
Science/Nature or was simply an unpublished ms.? I don't believe that for a
microtick! (And if it were remotely true, the RAE would be an even
blunter instrument than it is.)
It's the RAE's formal submission criteria that don't give a toss. But they are
just an empty formality, whereas push comes to shove once a submission is
(2) A fundamental point that has been trodden around but not mentioned
directly is the fact that the RAE (a) does not, and (b) cannot (for lack
of time and expertise) perform peer review on every submission. I am
certain that for all but a spot-checked minority sample the panel member
at most just dimly skims them. They rels instead on the expertise of the
peer review most of the (aricle) submissions have already undergone. That,
after all, is what peer review is for, and part of the function the
journal's name and track record are meant to perform.
Yes, of course Science and Nature also publish a lot of bilge. But, per saldo,
the Science and Nature papers are a good deal better than the unpublished mss.
And "per saldo" is what a monster exercise like the RAE is about. If it
were a serious individual peer evaluation of each submitted work it would
not only become a prohibitively time-consuming and expensive exercise
(calling for the consultation of worldwide experts, just as the peer
review did), but it would be monstrously profligate, for work that had
already been peer-reviewed once, properly.
But none of this speculation about how the RAE pundits go about their
bean-evaluation (when they could just as well be counting citations,
as Charles and others have shown) is pertinent to jisc-repositories and
its mission, is it?
Ours is just to deposit and die, not to post endlessly reasoning why...