Re: A study comparing 150 IQ+ persons to 180 IQ+ persons

From: Ben Goertzel (
Date: Mon Aug 21 2006 - 19:06:36 MDT


On 8/21/06, Michael Anissimov <> wrote:
> Abstract:
> "Although there are numerous studies of gifted students, there have
> been no studies following up the very high IQ students into adulthood.
> Using the Terman files, 26 subjects with scores above 180 IQ were
> compared with 26 randomly selected subjects from Terman's sample.
> Findings were generally that the extra IQ points made little
> difference and that extremely high IQ does not seem to indicate
> "genius" in the commonly understood sense of the word."

I read the article and the findings seem a lot more equivocal than the
abstract suggests... not anywhere near definitive...

> What to make of this? I'm not entirely certain, except to say that
> there is probably a point of diminishing returns with respect to
> intelligence in humans, after which other factors such as social
> skills, self-control, emotional stability, etc. begin to dominate over
> the effects of IQ.

What I make of this is very simple: The IQ test is a mediocre tool for
distinguishing levels of intelligence among different smart people.

I don't necessarily believe there is a point of diminishing returns
with respect to intelligence, though I do believe other factors (such
as persistence and self-understanding) are also highly important to
intellectual achievement.

But I do think that the IQ test is an obviously mediocre measure.

IQ measures, basically, the ability to solve moderately difficult
problems rapidly under immediate social pressure.

Whereas achieving great things intellectually is more about the
ability to solve EXTREMELY hard problems, without significant time
pressure or immediate social pressure. (Not to even mention the other
psychological aspects like persistence, self-understanding, etc.)

In computer science, algorithms that solve moderately hard problems
fast are not necessarily the same as ones that solve very hard
problems best given more time. Something similar probably holds for
individual humans' "cognitive algorithms."

According to this analysis, among the population of very bright
people, there is going to be a nontrivial correlation btw IQ and the
cognitive skills required for intellectual greatness -- but not a very
strong correlation. And this seems to be what the paper you
referenced suggests, rather than a total lack of correlation.

-- Ben

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:57 MDT