RE: Ornithoms, Parrots, and others (long)

["Stewart, Dwight" <Dwight.Stewart@VLSI.com>]

        -----Original Message-----
        From:   Dinogeorge@aol.com [SMTP:Dinogeorge@aol.com]
        Sent:   Thursday, November 12, 1998 12:40 AM
        To:     jjackson@interalpha.co.uk; th81@umail.umd.edu
        Cc:     dinosaur@usc.edu
        Subject:        Re: Ornithoms, Parrots, and others (long)

        In a message dated 11/11/98 10:45:49 PM EST,
jjackson@interalpha.co.uk writes:


        You can always run a different program on the same matrix and see
what turns
        up; or you can stick more characters in the matrix and run the
program again;
        or you can do both. If you're really lucky, your results will
reinforce one
        another, but usually they don't. Then you're back to square one, I
suppose.


        George's approach is a good idea.  You might wish to look for strong
interactions between
        Characters within the matrix, as well.  This occurs in nature all
the time, where 2 (or more) characters interact to give a result that is
different from any one character alone.
        I wonder if anyone has ever run a JMP (Microsoft data analysis
software) on data like this?
        I bet it would work or at least provide some interesting insights.

        Dwight