Re: Race, intelligence, and anti-racist prejudice (Was: Genetic Evolution)

L Olson (orion@sierra.net)
Mon, 27 Feb 95 01:18:30 GMT

ffujita@sun1.iusb.indiana.edu (Frank Fujita) wrote:
>First let me say that in general I think Arun Gupta has done a remarkable job
>of learning what he needs to know to make informed statements in this debate.
>
>Arun Gupta (gupta@mrspock.mt.att.com) wrote:
>: Clue to the
>: clueless : you must have read people commenting on the very tiny R^2
>: values. Find out what that means.
>
>One should be more charitable when pointing out the errors of others.
>If you want to find out why accounting for 10% of the variance is actually
>quite good, read:
>
> Rosenthal, R., & Rubin, D. B. (1982). A simple, general purpose
>display of magnitude of experimental effect. Journal of Educational
>Psychology, 74, 166-169.
> Ozer, D. J. (1985). Correlation and the Coefficent of Determination.
>Psychological Bulletin, 97, 307-315.
> Ahadi, S., & Diener, E. (1989). Multiple determinants and
>effect size. Journal of Personality and Social Psychology, 56, my
>preprint doesn't have the page numbers.
>
>I wouldn't want to call anybody clueless, since the understanding
>of the proper interpretation of r and R^2 requires more than clues.
>None of the above authors are even in the IQ debate, but the
>interpretation of r and R^2 is of interest to psychhologists who are
>not interested in IQ.
>
>Frank Fujita
>
>

Mr Fujita:

I'm glad to see a professional comment on this. I'd hope that my replies
to these fellows aren't literally taken as a slavish belief in everything
Herrnstein and Murray wrote - I'm needling them by taking the other
extreme. I've usually seen the opposition in this as defending an
assumption that racial and ethnic groupings in this country are perfectly
equal in their potential cognitive means, and that the measured
differences must be and are entirely due to racism and the socioeconomic
problems that spring from it. There's nothing especially wrong with that
assumption (other than simplicity), but it is an assumption.

Of course it isn't possible to identify some certain genetic contribution
from IQ data used in this book, and I don't think Herrnstein should have
made a guestimate on a subject like this. The principle, however, is
supported by some of the studies referenced. The elevated visuospatual
score is very consistent among both Asians and decendents in this country
and others. I'll post references if needed. Herrnstein gives three studies.
That isn't licence to misapply a principle to any racial grouping you
choose, but I don't see how the opposite case can be made then either. In
other words, and this was my point (I guess I've been too subtle about
this - it's been too much fun needling), remedial social engineering
programs, founded on the presumption that the mean cognitive potential is
equal among racial groupings, and so a presumption that equal results will
be obtained, aren't supported by anything we know or measure. They are
punches in the dark, and natural sources of divisivness, as they codify
race-based practices in society. Hiring for reasons other than ability is
also a drag on productivity - it literally makes us all poorer.

I have discussed this with my few psychologist acquantances, one of them
working in human factors and testing in aerospace. They generally agree
with Herrnstein (to my surprise really), and in fact this helped form my
view on it. I can post tables and equations if that's wanted, or discuss
r^2. I gave that short shrift because it seems to me that the opposition
wants bury the subject in reformulations of selected data as a diversion
(after failing to _redefine_ the sources of unsupportive data based on
Gould's book). Revisiting those calculations may be entirely valid, but is
typically used to cast bad light on a proponderance of good data. If I
need correction on this, I'm perfectly open.