How do you guys deal...

[The Mad Monk]

I don't see how you can maintain that test taking ability is a better measure of intelligence.

JM

If a man is physically strong but never does anything with his strength, is he still strong?

[MikeH]

I don't see how you can maintain that test taking ability is a better measure of intelligence.

JM

I still can't see any reason to use achievement as a measure of intelligence. There are so many factors that go into achievement and intelligence isn't even one of the factors in most cases.

[Jon Miller]

Because test taking isn't a measure of intelligence at high intelligence levels and certain great achievements are only possible if one is intelligent (we can probably forget business achievements for this, since capital is so important for those), although being in the right place at the right time and inspiration are also definite factors (inspiration might be included with intelligence?).

However, it seems to me that achievement is the only factor which can truly identify (not quantify, really) genius. You probably need to scale the number by some factor relating to the culture.

JM

[Lorizael]

If a man is physically strong but never does anything with his strength, is he still strong?

Not for long.

[Jon Miller]

If a man is physically strong but never does anything with his strength, is he still strong?

If your only measure of a person's strength is if he can pick up a 200 kg log, are you really measuring extremely high strength? Will you be able to identify the strongest people?

JM

[MikeH]

Because test taking isn't a measure of intelligence at high intelligence levels and certain great achievements are only possible if one is intelligent (we can probably forget business achievements for this, since capital is so important for those), although being in the right place at the right time and inspiration are also definite factors (inspiration might be included with intelligence?).

However, it seems to me that achievement is the only factor which can truly identify (not quantify, really) genius. You probably need to scale the number by some factor relating to the culture.

JM

Whether or not tests are a good measure of intelligence has no bearing on whether achievement is a good measure.

How do you rank an achievement on your genius scale? If you can't quantify something it's useless a measurement scale.

I think you are confusing "what criteria should we use to label someone a genius" with "what tool can we use to provide a quantifiable measurement scale for intelligence". Whatever that tool ends up being, it's hard to think of something that couldn't also be identified as a 'test' whether it be a written test, verbal test, or based on observation of other data about the individual.

Your point seems to be that that to earn the label of genius you need to have high intelligence, and to have used that high intelligence to create a work of genius. Which isn't unreasonable. But that isn't what we're talking about.

[Jon Miller]

And I think that that is the only true identifier of extremely high intelligence, which I can identify.

Yes, some people with extremely high intelligence will not be identified ( and some achievements are also not obvious, and so will not be identified).

That doesn't mean there is a better identifier.

We are looking for the best measure. If tests are not adequate, than we look for another. I agree that achievement does not have 1-1 correspondence to extreme intelligence, but it is the only way to identify extreme intelligence that I can determine.

JM

[MikeH]

So you've come up with a crap identifier which can only be used for a tiny minority of individuals, which you admit won't even identify that minority, and could even give you some false positives.

But for some reason you think it's still a superior measure than the entire concept of establishing some kind of generalised test that can be use for all levels of intelligence?

That is an... interesting choice.

[Jon Miller]

What evidence do you have that generalized testing works for those greater than 2 sigma above the average?

I work with datasets where I am looking for a very small number of events out of a very large number of events. It is very difficult to get good variables, and this is with simple physics processes! I can't imagine generalized testing is at all relevant beyond 3 or so sigma.

If I am looking for discovery of a small number of events, I will focus on things that give me a high purity rather than something which likely has little to know meaning.

JM
(And I don't think, if you remove business achievement, that 'genius achievement' will give many false positives. If you would like we could require multiple instances of genius achievement to increase the purity.)

[MikeH]

But you haven't proposed anything better. Your "identify acts of genius" concept is meaningless without a definition of what an act of genius entails. And you have already stated that even if you did use that as a measure it wouldn't identify all people of extreme high intelligence.

Does your work involve conversations like this:

"how are we going to identify this event?"

"we'll know it when we see it!"

Or conversations like this:

"we are struggling to identify extreme events because it's difficult to get good variables"

"ok we need to keep improving our detection processes so we can get better variables"

Current tests may or may not be good at identifying extremes of intelligence. Dunno. But at least it's a method which we can enhance and refine. You haven't proposed a method yet, just an abstract impossible to quantify concept.

[Jon Miller]

Our detection processes are limited. This is a fundamental problem of science.

We use the best variables we can, and identify 10% or 1% of our sample, if that is all we are able to positively identify.

In this case, I think it is quite obvious to say Mozart was a genius or Einstein (or Fermi or Feynman or Da Vinci). I didn't feel the need to go into genius achievements here, although I could of course. A beginning point would be 'achievements which would not be possible by people without extremely high intelligence'.

I don't think that the current tests are a method that can identify extremes of intelligence. I don't see how you can expect them to be able to. Most obviously, they are not created by people at the extremes of intelligence (actually, greater than the extremes of intelligence) and are not calibrated using a sample of extremely intelligent people (nor even highly intelligent people).

They are created and designed to classify people in the central 2 sigma of the distribution.

JM

[MikeH]

The data I need is this:

Scientists in the era of IQ tests who have had 'genius' achievements. Do they all have very high IQs?

If they do, it doesn't matter if the difference between a 170 and a 200 IQ is in the realms of statistical fuzz or not*, it's still a good predictor of ability to produce genius.

*and I don't know if it is, I haven't seen any evidence that it is or isn't.

[Al B. Sure!]

If a man is physically strong but never does anything with his strength, is he still strong?

How did this person become physically strong? He exhibited his strength in the process of building his strength (training).

IQ is inherent capacity; physically strength isn't.

[The Mad Monk]

Why are you arguing this? It's clearly a metaphor.

[Al B. Sure!]

Why are you arguing this? It's clearly a metaphor.

It's a bad metaphor. The difference between IQ and physical strength makes your analogy invalid.

If the badness of the metaphor is known by you, why even bring up such a bad comparison?

Never understood this dismissal when someone explains the invalidity of a metaphor... 'it's only a metaphor. shut up.' Like if your metaphor is meaningless, why did you even bring it up?

I don't understand people.