[SERIOUS] Clever fools: Why a high IQ doesn't mean you're smart

[Asher]

Correlation does not equal causation.

Perhaps I adhere to a stricter version of demonstrable than you.

[Kuciwalker]

Since when does "measure" imply causation?

Seriously, Asher, you're just being stupid at this point.

[KrazyHorse]

Asher, I suggest you think what you're saying through.

"Correlation is not causation" makes no sense in this context. Nobody is saying that Johnny takes an IQ test, does well on it, and then doing well causes him to be successful

IQ tests are a MEASUREMENT. All a measurement can do is to correlate with something you care about. The point is that this measurement actually does have some predictive power when it comes to real-world results.

[Felch]

Who even takes IQ tests anymore? I've been subjected to a wide array of testing, both for school and as a volunteer in psychological studies, but nothing has ever explicitly quantified my "IQ". My chief experience has been seeing them in ads for who knows what. There's even an ad about counting triangles on my screen right now!

Has anybody actually taken one?

[KrazyHorse]

I took two formal IQ tests and one test which was IQ in all but name. One was at 4 years old which my mother paid to have done. One was when I was 5, before they put me up a grade. One was at the age of 9-10, which again my mother paid to have done.

My mother has a BSc in psych, so she's interested in this type of stuff. She's a big believer in testing, g, the whole deal.

[Asher]

Asher, I suggest you think what you're saying through.

"Correlation is not causation" makes no sense in this context. Nobody is saying that Johnny takes an IQ test, does well on it, and then doing well causes him to be successful

IQ tests are a MEASUREMENT. All a measurement can do is to correlate with something you care about. The point is that this measurement actually does have some predictive power when it comes to real-world results.

That's true enough. I still don't think testing well on an IQ test necessarily correlates well with intelligence in many people. I've spent too much time around arrogant ****s who scored sky-high on IQ tests only to fail at the simplest of tasks in the real world.

I think scoring well on an IQ test correlates well with success in our society, which means the people who take tests well and test well on things such as the IQ test will do well to become great business managers who make monumentally stupid decisions on a daily basis.

I suppose my objection is more to the point that high IQ test scores mean you're intelligent. I don't think that's the case. There probably is some correlation, but from my experience it's nowhere near enough to be a general rule.

[rah]

Back in the 60's it was quite common and almost all young students took at least one and it was used by aptitude specialists that would pretend to predict what your best career path should be. It was quite entertaining but some parents took it way to seriously.

So I'd assume that anyone over 50 here has taken a standardized version of one.

[Lorizael]

There probably is some correlation, but from my experience it's nowhere near enough to be a general rule.

The problem here is that you have some internal sense of what human intelligence is which you believe does not correlate with IQ. There needs to be a good definition of what exactly constitutes human intelligence before we can have a good test of what measures human intelligence.

[Asher]

Who even takes IQ tests anymore? I've been subjected to a wide array of testing, both for school and as a volunteer in psychological studies, but nothing has ever explicitly quantified my "IQ". My chief experience has been seeing them in ads for who knows what. There's even an ad about counting triangles on my screen right now!

Has anybody actually taken one?

I've only taken an informal one out of curiosity, and scored well. I'm not sure if it's the exact same as a "professional" grade IQ test, but it was pretty epicly stupid.

It's been my experience that the people who take IQ tests tend to have parents who want to think their baby is an Einstein, or people who want to think themselves to be one. The success correlation with high IQ tests is probably more related to the drive and circumstances of people who take such tests more than anything else.

[Asher]

The problem here is that you have some internal sense of what human intelligence is which you believe does not correlate with IQ. There needs to be a good definition of what exactly constitutes human intelligence before we can have a good test of what measures human intelligence.

I've taken classes with PhDs in mathematics who would undoubtedly score sky-high on IQ tests who has never figured out how to properly be sociable with people, or to be able to troubleshoot basic problems in life. A couple of the kids I knew in high school who have always bragged about high IQ scores -- and tested well in school in general -- do and say stupid things all the time. One such guy took a MA in Philosophy and has only now tried to consider what kind of job he can get with the degree, he never thought ahead (and yes, this is just coincidence with Boris...)

I consider such actions to be pretty stupid. But I doubt those are tested on IQ tests...

[Asher]

Another example is someone I know on this forum who probably scored high on an IQ test, but is looking for jobs in and around Washington DC and Pittsburgh.

Like, WTF?

[KrazyHorse]

Asher, I'm agnostic about the meaning of the word "intelligence". I care more about correlations, predictive power, modeling, that sort of thing.

Also, you should bear in mind that high end IQs are basically bunk; we don't have good statistics on correlation at the high end. If you want to argue that the predictive power of IQ tests tails off past, say, 2 sigma then I don't know if I have enough evidence to argue with you. There are some professions in which the correlation continues to higher than this (physics is most certainly one; I don't know if any of the profs in this dept. would test at less than 3 sigma, and many/most would test at 4 sigma at a guess).

The point is simply that IQ measures SOMETHING that's not totally meaningless. Of course it's not a perfect tool.

[Asher]

It measures the ability to take analytical tests more than anything else. Society values that.

[Felch]

I took a bunch of tests with the Maryland Psychiatric Research Center and Johns Hopkins for a couple different NIH studies. Lots of multiple choice pattern matching, short term memory, word recognition, and some stuff like recognizing emotions or recognizing a face from different angles and lighting.

The best was a little game where there'd be a pair of pictures on the screen. You had to figure out which one was the correct choice, but the rules were never revealed and they kept changing at random. So for the first twenty it might be pick the biggest, for the next six it could be pick the shorter image, and then the darker or heavier or whatever. You were paid 5 cents for each correct answer, deducted 5 for every wrong answer and penalized 10 cents for taking too long (you had three seconds or so per pair of images). I wound up making a few dollars, but it took a while.

They never told me a score, but I had fun. And I was paid 15 bucks an hour to do it.

[KrazyHorse]

at felch skewing the results of psych tests...