Alexander Dekhtyar wrote:
>
> In order for it to work, there really has to be a high probability that
> the reviewer is not familiar with the previous work of the authors on
> similar topics. This is achievable in one of two situations:
>
> - large communities;
>
> - smaller communities that use "outsiders" as reviewers.
>
Reviewers who are "outsiders" to the work concerned (whether they come
from outside a small community, or from within another group in a large
community) may be less likely to see the originality or significance of
very orginal or significant work. On the other hand, they may be more
likely to view as original or significant work which is not either. In
both cases this is because such outside reviewers won't know or fully
appreciate the tradition from which the work arises.
Given that there may be many such tendencies operating both in favour
and against authors, my conclusion is that the only method fair in the
long term is to make everything open to debate, i.e.,
subject to contestation and defence. This says to me that reviewing
should be double-sighted (the opposite of double-blind) with both
authors and reviewers putting their names to whatever they write.
But I presume this makes the task of Journal Editors or Program
Committee Chairpersons harder, as they cannot choose just anybody to
review work. They too may have to justify their choices. Is this a
reason why double-sighted reviewing is not widespread?
-- Peter McBurney
****************************************************************
Peter McBurney
Agent Applications, Research and Technology (Agent ART)
Group
Department of Computer Science
University of Liverpool
Liverpool L69 7ZF
U.K.
Tel: + 44 151 794 6768
Email: P.J.McBurney@csc.liv.ac.uk
Web page: www.csc.liv.ac.uk/~peter/
****************************************************************
This archive was generated by hypermail 2b29 : Mon Nov 19 2001 - 10:29:23 PST