Re: [UAI] UAI and double blind reviews

From: Subbarao Kambhampati (rao@asu.edu)
Date: Wed Nov 14 2001 - 18:38:21 PST

  • Next message: Alexander Dekhtyar: "Re: [UAI] UAI and double blind reviews"

    Just as another view to counter-balance Robert's, let me say that I too
    have seen double-blind reviewing
      in use in AI conferences and found it to be a net plus. I did not find
    it all that hard to write without intentionally divulging my identity, and
    neither did I find a compelling need to
    know the identity of the author of the paper I was reviewing.

    (For the record, I have written as well as reviewed a fair number of papers
    of each type. Also, for the record, the discussion is purely academic for
    me since I don't normally send anything to UAI.)

    I also had a chance to experience first hand the "perceived" advantages
    that double-blind reviewing brings to an outsider trying to get "in", when
    SIGMOD--the premier database conference--turned to double blind reviewing
    starting last year. It was tried on an experimental basis in 2000, and has
    now become the default. Other database conferences are expected to follow
    suit. (To my knowledge, there is no conference that tried double-blind
    reviewing and abandoned it. There is always the usual initial resistance to
    the idea, but once tried, it seems to stick...)

    I think double blind reviewing serves the same kind of purpose as the
    strictures on the jurors in a jury trial against discussing the case
    outside the court room--it doesn't worsen the quality of reviewing much,
    and it can improve the *perception of fairness* of reviewing significantly.

    Rao
    http://rakaposhi.eas.asu.edu/rao.html

    At 11:20 AM 11/14/2001 -0800, Robert P. Goldman wrote:

    >Well, if this issue is going to be bruited about on this forum, I'd
    >like to cast my vote AGAINST the double-blind process. I've seen that
    >at work in the AI conferences, and it's a monumental pain to deal
    >with, both as a reviewer (where I have to pretend I don't know people
    >whose work I recognize perfectly well) and as a submitter, where I
    >have to try to launder my references. And as a reviewer, I find
    >ENORMOUS differences in the degree to which submitters comply with the
    >anonymizing instructions.
    >
    >Personally, I will be delighted to keep an appearance of greater
    >amateurism in UAI. And I don't think we need to comply with the
    >practices of any organization that would site its conference in
    >Detroit :-)
    >
    >R



    This archive was generated by hypermail 2b29 : Wed Nov 14 2001 - 18:41:15 PST