Path: utzoo!utgpu!water!watmath!watdragon!violet!wphughes
From: wphughes@violet.waterloo.edu (William Hughes)
Newsgroups: sci.misc
Subject: Re: Bias on IQ tests
Summary: IQ tests are not meaningless
Message-ID: <6041@watdragon.waterloo.edu>
Date: 28 Mar 88 22:16:30 GMT
References: <153reneerb@byuvax.bitnet> <20779@bu-cs.BU.EDU> <3933@bloom-beacon.MIT.EDU> <3943@bloom-beacon.MIT.EDU> <4087@bloom-beacon.MIT.EDU>
Sender: daemon@watdragon.waterloo.edu
Reply-To: wphughes@violet.waterloo.edu (William Hughes)
Organization: U of Waterloo
Lines: 41

In article <4087@bloom-beacon.MIT.EDU> tada@athena.mit.edu (Michael Zehr) writes:
>Well, I once had an analyst tell me that he and a lot of other psychologists
>believed that the score you get on an IQ test has about as much to do with
>how well you succeed at life (college, and other things that are supposed
>to require intelligence, in particular) as the length of [insert favorite
>euphimism for male sex organ here].
>
>Furthermore, results from IQ tests can change redically from year to year,
>so that makes them even more suspect.

The new popular mythos, that IQ test are meaningless, is almost as
inaccurate as the old popular mythos, that IQ tests are a single, almost
perfect, measure of innate intelligence.  Despite the beliefs of your
analyst, IQ correlates quite highly with scholastic success.  The correlations
with more general success are not as marked but still quite significant.
So, whatever the IQ score is, it is clearly of some interest.

True scores do change, and sometimes radically, but on the average not
enough to invalidate the use of the instrument.  Test/retest correlations
have been done (the researchers in this field are not all idiots!)
and are relatively high.  However, it is true that the inaccuracies
of the test make its use for individual evaluation highly problematic
(to say the least).  Its use as a research tool by those who understand
its strengths and limitations is entirely justified.  I very much like
Craig Werner's benchmark analogy.

Check out some of the copious literature on the subject if you are interested.
(Be careful some of it is crap!)  For a very good, readable, well balanced
review of intelligence testing see "Inteligence: nature, determinants and
consequences" by E. B. Brodie and N. Brodie, Academic Press, 1976.

It should be noted that in "Mismeasure of Man" Gould does not argue that
IQ tests are meaningless.  He does argue that they have been radically
overinterpreted by many scientists in an effort to perpetuate
stongly held biases.  He also argues that the tests are not in themselves
sufficient to establish the hypothesis that there is a single entity
called "generalized intelligence" which can be measured.  It is a very good
book, but it is not the whole story on intelligence testing nor does Gould
claim that it is.

                                           -William Hughes