Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!cs.utexas.edu!uunet!mcvax!ukc!castle!aipna!edai!cam
From: cam@edai.ed.ac.uk (Chris Malcolm    cam@uk.ac.ed.edai   031 667 1011 x2550)
Newsgroups: comp.ai
Subject: Re: Inheritance of IQ
Message-ID: <496@edai.ed.ac.uk>
Date: 5 Aug 89 01:50:06 GMT
References: <5453@pt.cs.cmu.edu> <2061@cbnewsh.ATT.COM> <5480@pt.cs.cmu.edu> <458@edai.ed.ac.uk> <602@visdc.UUCP> <3072@cadre.dsl.PITTSBURGH.EDU> <603@visdc.UUCP> <482@edai.ed.ac.uk> <611@visdc.UUCP>
Reply-To: cam@edai (Chris Malcolm)
Organization: University of Edinburgh, Edinburgh
Lines: 39

In article <611@visdc.UUCP> jiii@visdc.UUCP (John E Van Deusen III) writes:
>In article <482@edai.ed.ac.uk> cam@edai (Chris Malcolm) writes:
>>In article <603@visdc.UUCP> jiii@visdc.UUCP I wrote:

>>> The levels of genetic intelligence within a population fit a standard
>>> probability distribution; that is, 95% of the population is within
>>> two standard deviations of the mean.  
>>
>> This is tautologous. It was PRESUMED by those constructing IQ tests
>> that the results OUGHT to fit a normal distribution; so they fixed the
>> test so that they did.  There is no other reason why the distribution
>> should be normal: many simple physical measurements have skewed normal
>> distributions, bipolar distributions, etc..
>
>Whether a characteristic such as the "level of genetic intelligence" can
>be measured is irrelevant; we have a pretty good idea that it exists.
>If it could be measured reliably, being a slowly varying characteristic
>of 5 billion people, it would probably fit a normal distribution.  Do
>you really have any evidence that they "fixed" the test?

Nothing to do with "fixing" - check any textbook on the construction of
IQ tests - it's the way it's done. Like you, they presume that "it is a
slowly varying characterisitc ... which would probably fit a normal
distribution". But in a social creature like man, I find it just as
plausible that the best evolutionary compromise might be an uneven
bipolar distribution, which would tend to provide the average tribe size
with at least one smart guy to solve the really difficult problems, but
not so many as to make the tribe unmanageable by the leadership. I'm not
arguing this - I merely want to point out that we simply don't yet know
anything about the distribution of intelligence in human populations;
nor do we have a clue as to whether the average Nobel prizewinner is
twice as smart as the man in the street, or two hundred times as smart.
IQ is not a linear scale, it's simply a ranking order mapped onto
numbers and massaged statistically in a vain and foolish case of physics
envy.
-- 
Chris Malcolm    cam@uk.ac.ed.edai   031 667 1011 x2550
Department of Artificial Intelligence, Edinburgh University
5 Forrest Hill, Edinburgh, EH1 2QL, UK