THE BELL CURVE and PCT argument on statistics

#^#^#^#^#^#^#^#^# FROM CHUCK TUCKER 941201 #^#^#^#^#^#^#^#^#

        I have been using the PCT argument on use of statistics
        in the behavioral sciences and today read Stephen Jay
        Gould's review of the book in THE NEW YORKER 941128
        which states: "Herrnstein and Murray's correlation
        coefficients are generally low enough by themselves
        to inspire lack of confidence. .... Although low figures
        are not atypical for large social-science surveys involving
        many variables, most of H and M's correlation are very
        weak - often in the 0.2 to 0.4 range. Now, 0.4 may
        sound respectably strong, but - and this is the key point -
        R2 is the squarre of the correlation coefficient, and the
        square of a number between zero and one is less that the
        number itself, so a 0.4 correlation yields an r-squared
        of only .16. In Appendix 4, then, one discovers that the
        vast majority of the conventional measures of R2, excluded
        from the main body of the text, are less that 0.1. These
        very low values of R2 expose the true weakness, in any
        meaningful vernacular sense, of nearly all the relationships
        that form the meat of 'The Bell Curve.'" (p. 147) Sound
        familiar to anyone!

        BTW, Newsweek in their report on this book interpreted R
        as R2 and said something like ".4 correlation accounts
        for 40% of the variation" - I was amazed.

        On Modelling

        I would really appreciate it if someday one of you modellers
        could provide some instructions for using these model
        program for teaching and perhaps research. I think it would
        a nice topic for an essay - "How to use ..... "

        Regards, Chuck