I don't know about "IQ tests seem to be terrible at measuring actual brilliance." - Some of the components of IQ tests are fantastically accurate. I can't read maps, I can't put a monitor back in it's box. And I score very poorly on the pattern assessment components of IQ tests. I guess you might say I'm "mid-to-high level mentally retarded on pattern assessment."

Likewise, I have a friend who is world class _stellar_ at putting Monitors back in complicated styrofoam packing and into boxes (Seriously - his ability in this task is amazing) - and he tests off the chart on the pattern recognition component of IQ tests.

So - I can, based on the admittedly small anecdotal sample of n=2, suggest that there is strong predictive power of at least some elements of the IQ tests.

I think the main point is that IQ measures some sort of current fluid intelligence. It doesn't measure other types of intelligence, and most importantly it doesn't measure potential. So the blanket concept of it defining someone's common-understanding-of-"intelligence" is flawed to the extent of it being a pointless or harmful measure in many of cases.

What I'm arguing is the opposite - that the IQ test actually is precisely a measure of potential in various abstract areas. Whether you do anything with that potential, or how you develop it (as, in the case of my friend, who is a savant in putting Monitors back into styrofoam boxes) - is a function of resources, opportunity, and interest.

Flynn goes further, and suggests that the society and culture that you grow up in, also enhance your ability to execute on IQ tests. The genetic potential (obviously?) hasn't changed, but the environment has shifted such that our "actuated" IQ has increased as we practice more often in abstract tests.

>> that the IQ test actually is precisely a measure of potential in various abstract areas.

I'm trying to understand this. Here's my understanding: That it is possible to take an IQ test and have score x, then perform activities y, retake the IQ test and have the score x+15. How exactly did the initial test measure the potential for that increase? That your potential is x (+-)15 (or 25 or whatever the number is)?

IQ measures your potential to perform and learn in various areas. Per wikipedia, one of the standard suite of tests includes, "Comprehension-Knowledge, Long-Term Retrieval, Visual-Spatial Thinking, Auditory Processing, Fluid Reasoning, Processing Speed, Short-Term Memory and Quantitative Knowledge and Reading-Writing Ability"

The argument that James Flynn makes, is that your potential can increase, through your environment.

That is - your IQ isn't simply a function of your genetics, but also of what you are exposed to.

To put it in concrete terms - someone who worked every day, from dawn to dusk, in very manual labor - let's say cutting wood, chopping it up, and stacking it, will develop their potential for abstract thinking very differently than if they had instead gone and gotten their PHD in some abstract field like Logic.

Same person, same genetics - very different IQ.

What you are referring to, is closer to the "Potential for Potential" - I don't know if there is any way, or anyone who measures something like that. But I think that everyone will agree, that IQ is _not_ a measure of your "Potential for Potential" - simply a measurement of your potential - and that is malleable.

I think many people believe that IQ is your potential for potential, and indeed the potential for potential of your descendants and your "race". Thats what people generally mean when they say Group X has a high/low IQ. I think this might be referred to as the strong IQ hypothesis.

> What you are referring to, is closer to the "Potential for Potential" - I don't know if there is any way, or anyone who measures something like that.

Your parent poster is definitely referring to this concept. (S)he is referring to that concept explicitly because it is so ill-definable and unmeasurable.

It's pointless to judge an accomplished person, like a businessman or scientist, with an IQ test, because that misses the point. It's just a very quick (15 minute) way to categorize people efficiently.

So, if the army wants to determine who should be an officer candidate, or an NCO, or a radar operator, or a potato peeler, when they have 3 million men to process, it's incredibly efficient. If you are interviewing 10 candidates for a management position, an IQ test would make no sense.

Why does it have to be about brilliance. The US armed forces is one of the bigger users of IQ tests and has plenty of data showing higher IQ scores leads to better performance.

I think it would be fascinating to view the outliers (low IQ when measured=high performance) and investigate the reasons.

(My hypothesis:) Naturally what I would expect would be found is that IQ is just one measure that sometimes correlates in certain types of people, and that it completely fails and miscalculates actual intelligence and mis-predicts future performance in other types of people. And that it fails in a statistically significant number of cases. And that due to the behavior confirmation effect (self-fulfilling prophecy) it ends up true in more cases than it could.

Meaning IQ is an OK measure in many cases, but very flawed. I'm curious what will eventually take it's place as more useful.

It is my understanding the DoD uses IQ test only for initial entry into the various services. I believe the threshold is 105 (although some military guys have said that at the peak of the Iraq & Afghan campaigns recruiters would let a few sub 105s in, quotas & bonuses.) I would be interested in a link to those studies (if you have them available.)