On re-doing analysis

From: Robert A. Knop Jr. (robert.a.knop@vanderbilt.edu)
Date: Tue Feb 25 2003 - 10:44:05 PST

  • Next message: Tony Spadafora: "Re: On re-doing analysis"

    Several people have called for re-doing the analysis based on this,
    that, or the other thing. Some say, (e.g. Greg on the Jha U-band stuff)
    that "this won't cause much of a delay".

    I beg to differ. This will cause a HUGE delay. Doing all of these fits
    and so forth takes time. (Training somebody else to do it will take
    just as much time, so that's not a real solution.) Depending on how
    much we want to do, we should be aware that we're potentially talking
    about a lot of time. Putting in a different U-B intrinsic color means a
    solid week of work on my part-- and that's a real week, not a week
    interupted by meetings and the other things that happen. That's pretty
    early in the analysis chain (i.e. whole new uberspectrum, probably a
    whole new U-band template), and must be propogated all the way through.
    I can guarantee you that the next draft of the paper will not come out
    until April, at the earliest, if we insist on making this change. We're
    not talking about plugging in a small change for a number quoted in the
    paper, we're talking about redoing everything here.

    Is Jha's work published in a refereed journal, or is it just available
    as his thesis? If the latter, I think we are justified in ignoring it
    for now. If it is published, where and when?

    Another issue: WMAP. So far as I can tell, that work is not published
    either, just submitted. I used that as a justification not to mention
    what they'd done in the last draft Doing a lot with that -- even a lot
    of discussion -- will take time and delay the paper. If their paper
    isn't published, only submitted, we can make the argument that this is a
    "simulatenous" paper, or that their results weren't available when the
    paper was written. Reason: to get this paper out the door. That should
    be the main goal. A perfect paper will never be published; by the time
    we deal with all of that, more stuff will have come out that we will
    have to deal with, and we'll be stuck like Xeno's paradox never reaching
    the finish line. I would rather get this one out the door, and then
    write a separate paper that combines these results with WMAP results.

    I understand that we want to do this right. We also want to get it out
    the door. Even if there is something better around the corner, I'd
    rather we publish than we keep waiting to plug in the next best thing we
    see coming. I realize this wouldn't be an issue if I'd gotten the
    bloody draft written last August, but, well, that didn't happen.

    -Rob

    -- 
    --Prof. Robert Knop
      Department of Physics & Astronomy, Vanderbilt University
      robert.a.knop@vanderbilt.edu
    


    This archive was generated by hypermail 2.1.4 : Tue Feb 25 2003 - 10:44:06 PST