Is IQ real?

(Or does it default to integer?)*

Jordan Peterson has some comments here.

From the autogenerated transcript:

1:11 one of the things I have to tell you
01:13 about it IQ research is that if you
01:15 don’t buy IQ research you might as well
01:19 throw away all the rest of psychology
01:21 and the reason for that is that the
01:24 psychologists first of all who developed
01:26 intelligence testing were among the
01:28 early psychologists who instantiated the
01:30 statistical techniques that all
01:32 psychologists use to verify and test all
01:35 of their hypotheses so you end up
01:37 throwing the baby out with the bathwater
01:39 and the IQ people have defined
01:42 intelligence in a more stringent
01:44 stringent and accurate way than we’ve
01:47 been able to define almost any other
01:50 psychological construct and so if you
01:52 toss out the one that’s most well
01:54 defined then you’re kind of stuck with
01:55 the problem of what are you going to do
01:57 with all the ones that you have left
01:58 over that are nowhere near as
01:59 well-defined
02:00 or as well measured or as or as or or
02:04 whose predictive validity is much less
02:07 and has been demonstrated with much less
02:09 vigor and clarity

Also here:

00:01 so IQ is reliable invalid [and valid – ed] it’s more
00:05 reliable and valid than any other
00:06 psychometric test ever designed by
00:09 social scientists the IQ claims are more
00:11 psychometrically rigorous than any other
00:13 phenomena phenomenon that’s been
00:16 discovered by social scientists

Also of interest:

08:32 should tell you how to make an IQ test
08:33 is actually really easy and you need to
08:36 know this to actually understand what IQ
08:38 is so imagine that you generated a set
08:42 of 10,000 questions okay about anything
08:45 it could be math problems they could be
08:47 general knowledge they could be
08:49 vocabulary they could be multiple choice
08:50 it really doesn’t matter what they’re
08:52 about as long as they require abstract
08:53 to solve so they’d be formulated
08:56 linguistically but mathematically would
08:58 also apply and then you have those
09:01 10,000 questions now you take a random
09:03 set of a hundred of those questions and
09:05 you give them to a thousand people and
09:08 all you do is sum up the answers right
09:10 from so some people are gonna get most
09:12 of them right and some some of them are
09:13 going to get most of them wrong you just
09:14 rank order the people in terms of their
09:16 score correct that for age and you have
09:19 IQ that’s all there is to it and what
09:22 you’ll find is that no matter which
09:24 random set of a hundred questions you
09:26 take the people at the top of one random
09:28 set will be at the top of all the others
09:30 with very very very high consistency so
09:34 one thing you need to know is that if
09:36 any social science claims whatsoever are
09:39 correct then the IQ claims are correct
09:44 because the IQ claims are more
09:46 psychometrically rigorous than any other
09:48 phenomena phenomenon that’s been
09:51 discovered by social scientists

*  Fortran reference

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s