Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!sdcsvax!dcdwest!ittvax!decvax!cca!ima!inmet!lutton
From: lutton@inmet.UUCP
Newsgroups: net.misc
Subject: Re:  IQ tests
Message-ID: <1689@inmet.UUCP>
Date: Thu, 4-Oct-84 00:33:43 EDT
Article-I.D.: inmet.1689
Posted: Thu Oct  4 00:33:43 1984
Date-Received: Sun, 30-Sep-84 04:18:19 EDT
Lines: 22
Nf-ID: #N:inmet:6400119:000:959
Nf-From: inmet!lutton    Sep 26 12:42:00 1984

<>
I've mailed a note to the person who asked about IQ statistics, but
in case anyone is interested, here is some info.

IQ tries to measure linearly something that is multi-dimensional.
It's all pretty silly, but its use continues for lack of anything
better.  Binet's original tests were to be used for placing little
kids in the proper classes.

IQ = mental age / chronological age.  People don't get smarter after
age 16 so chronological age for adults is set to 16.  Lots of people
are tested, and mental age is set so that IQ equals 100 for the mean.
IQ follows a normal distribution, so 100 is also median and mode.

Standard deviation is 16 IQ points.  One person in 10,000 has an
IQ of 164 or better -- 4 standard deviations from mean.  I guess
that means one in 10,000 has 36 or worse.

If you take multiple IQ tests the results will vary over a wide range.
If you take enough, you will score 138 or better sooner or later.
Then you can join Mensa.