Further Evidence That IQ Does Not Measure Intelligence | Hacker News Reader

Further Evidence That IQ Does Not Measure Intelligence

io9.com

19 pointsColinWright66 comments

In my own life, I see more and more "evidence" that IQ, or academic achievement, is the wrong metric for intelligence. When I grew up, everyone was telling me that intelligence is in direct relation with how you perform at school. Someone who performs well at school is by definition intelligent, someone who is not good at school is not intelligent. Simple.

This went on after high school. Here in Belgium, the better students go to university while the less achieving students go to something directly translated to English as "high school". A "high school" in Belgium is a more applied type of study. For instance, you won't see any theoretical computer science course (decidability, complexity theory) taught at a "high school". Again, it was assumed that universities produce the intelligent people and "high school" the less intelligent people.

During the past few months, being part of a startup, I've worked with several people who in my eyes are intelligent. These guys can come up with creative solutions to problems, well into the realm of out-of-the-box thinking. However, they don't all come from universities. Some have a "high school" degree, or no degree at all. It made me realise that academic achievement is not a good measure for creativity. While for me, creativity is an important factor in intelligence. Many fellow students in my university computer science course were "intelligent" by academic achievement (after all they were able to finish an engineering university degree) but I wouldn't call them intelligent, just because they lacked this creativity. They were good when they were being told what to do but had a hard time thinking out-of-the-box. And I think that creativity is so important in the definition of intelligence.

I'm not saying that intelligence is all about creativity. Someone who can think creatively and think analytically is more intelligent than someone who is just good at thinking creatively. But I'm saying that academic achievement (typically related to IQ) is not a good measure for intelligence because it lacks that "creativity" factor. "Ken Robinson says school kills creativity" is a great TED talk about this (http://www.ted.com/talks/ken_robinson_says_schools_kill_crea...).

> IQ, or academic achievement

> When I grew up, everyone was telling me that intelligence is in direct relation with how you perform at school.

Your mistake is equating IQ with academic achievement. That's not what it is. It helps. But it's neither necessary nor sufficient. Academic achievement is something else.

IQ, or intelligence quotient, is exactly supposed to be a measure of intelligence (if intelligence is at all measurable). Academic achievement isn't supposed to be related, as ars said in a sibling post (although IMHO there is some correlation between the two).

"You Are Probably Not Much Smarter or Dumber Than Anybody Else"

I've noticed that people seem to be completely unable to really understand that some people are actually smarter than them. Most commonly they will confuse aptitude for intelligence. i.e. they easily acknowledge that this person just likes intellectual activity. But can't seem to grasp the difference in intelligence.

The same [relative] lack of intelligence also robs people of the ability to understand intelligence in others.

It's easy to see when someone is less intelligent - it's much harder to really grasp that someone people are smarter. It's like trying to explain a red to a colorblind person who can't see red. They kinda get that it's another color like others they've seen before, but they just can't grasp that it really exists.

You also (less commonly) get the reverse - someone really smart doesn't understand why everyone can't do it.

An example: Srinivasa Ramanujan heard the number 1729 and instantly thought of an interesting property for it. This ability to do that for a number just doesn't exist in most people. It doesn't matter how hard they try or study - they just can't do that.

1*7+2=9

That's not a property of the number 1729, that's only a property of the decimal notation for the number; an important distinction. The property is also trivial. Ramanujan's property - that the number is expressible in two different ways as a sum of two cubes, and in fact the smallest number so expressible - is a property of the number. That Ramanujan had different innate abilities from the vast majority of people is about as clear as the midday sun.

There are two problems though with this view. As a disclaimer, there are people who I consider much smarter than I am at certain things, and many people regard me the same way. Some of that may be my attention deficit and the effort I have had to put in to being successful despite it. I think the problem though is different, however. I am going to say two things:

1. Intelligence is not definable to the point where it can be tested, and

2. Attitude > innate intelligence.

I expect both these will not be so welcome in the HN community as such, so I am prepared for downvotes, but I also dont think these are wrong either.

On the first, the fact is that people process information differently. Ramanujan processes a number a certain way. Someone else might associate it with a year and immediately make historical connections. Someone else may factor get fascinated and factor it. There are bunches of ways a piece of information can be processed and the weighing of one as better than another is ultimately a cultural judgement and ultimately what proves to be an interesting property depends on what we find fascinating (which brings me to my second point).

On the second, I have found all my life that the smartest people I have known have always been the most humble, and to regard everyone else as being worthy of learning from.

I have consequently come to regard intelligence as a two-fold habit. The first part is the habit of finding everything interesting in life, what I call an unwillingness to be bored (something that really comes through in Feynman's memoires). The second is regarding everyone as someone to learn from (Heisenberg btw comes across this way very much in "Physics and Philosophy"). If you go around learning from everyone, and refuse to be bored, you will quite quickly become smarter than everyone else if they refuse to do the same.

>> 2. Attitude > innate intelligence.

What do you mean by greater than?

Attitude like you describe is only really open to people above a certain level of intelligence. If you just don't get mathematics, language etc etc because of inbuilt limitations (yes, some people have these) then you're not going to get that far.

What I mean is that most of intelligence intelligence is a habit cultivated by attitude rather than limited by an inborn capacity. This doesn't mean there isn't a basic shape given by innate aspects (at least for some people), but that most of what we'd call intelligence is developed rather than inborn.

I would be interested to see any research you have seen that points in that direction. I think that people are more restricted by hardware capabilities than you like to imagine.

That said, I asked what you meant by 'greater than' because I don't believe that raw intelligence always translates to success at life, whatever that may mean.

Well, at some point software affects hardware though regarding our brain.

Could Sapir or Chomsky have been great linguists if they only had one native language and learned the other languages at, say, the age of 10?

That's true, there is definitely environmental feedback involved.

However people can grow up in the same environment, with the same access to learning, and have very different outcomes.

You are assuming that most people fulfil their potential.

As I mentioned on another post the way I look at it is that we are granted a "fated" potential that we must strive to reach. I don't think most people do this effectively.

>>You are assuming that most people fulfil their potential.

No, I'm really not, I'm saying that people have different levels of intelligence and you can see this as they come out differently from the same environment.

You seem to have a rosy view of everybody as smart and equally capable if they just jolly-well gave it a go. This is just not the case.

> You seem to have a rosy view of everybody as smart and equally capable if they just jolly-well gave it a go. This is just not the case.

No, I just think most people could be quite a bit more smart than they are. That does not imply equal potential.

Wikipedia cites numerous studies on heritability, showing that IQ starts off largely environmental (0.2 heritability at birth) but ends up eventually moving to what its been genetically predetermined to be (0.8 heritability in adulthood).[1] IQ seems to accurately measure abstract reasoning abilities, pattern recognition abilities, and memory retention. That's what most people think of when they say intelligence (though there are always efforts to "redefine" intelligence to make it a more politically correct / egalitarian term - at which point, you're not really talking about "intelligence" but rather "semantics" and how you want to define a word in the English lexicon).

The IQ "debate" reminds me very much of the race debate, that race "doesn't exist" and is purely a social construct with no genetic basis. I keep waiting for someone to take a black or Asian baby and place it with white parents for it to grow up and have white skin and a pointy nose and blond hair. So far I haven't seen it.

We have no trouble understanding that the differences between, say, a sheep and a cow are due to genetics; why is it so hard for us to wrap our minds around the fact that most of the differences between one another come down to genes, too?

Methinks it comes down to the old philosophical debate of free will vs. predetermination, and the innate fear many people have of the idea that you can do anything or be anything if you only work at it hard enough.

[1] https://en.wikipedia.org/wiki/Intelligence_quotient#Heritabi...

> Methinks it comes down to the old philosophical debate of free will vs. predetermination, and the innate fear many people have of the idea that you can do anything or be anything if you only work at it hard enough.

For the record, I don't think it works that way either. I am largely a fan of the idea that F. M. Cornford reconstructed from Greek myth ("From Religion to Philosophy"), that we have free will within a granted partition of fate, or lot of life, which is limited in various ways and repaid when we die. From this view, fate is a domain we must strive to achieve, and thus not inevitable.

> The IQ "debate" reminds me very much of the race debate, that race "doesn't exist" and is purely a social construct with no genetic basis. I keep waiting for someone to take a black or Asian baby and place it with white parents for it to grow up and have white skin and a pointy nose and blond hair. So far I haven't seen it.

Race is problematic in a number of levels. Yes it is a social construct with no real genetic basis, at least as we have constructed it. This doesn't mean however that you can't identify genetic commonalities among different groups. We could, for example, just as easily say that the racial test is drinking milk as an adult, and therefore Mongolians and Scandinavians are of the Milk-Drinking Race, while Italians and Chinese are of the Cheese-Only race. And before you say "but skin is more visible" it is worth noting that eating together and food taboos are historically much closer to ethnicity than skin color is.

The problem though I think isn't a question of mere egalitarianism. There are people who are for all intents and purposes, greater visionaries, better at recognizing patterns that will change the world, and so forth than others. IQ tests, IMO will never be able to capture that for very fundamental epistemological reasons.

This isn't to say that IQ isn't useful but as long as you are weighing different skills against eachother, assuming the only pattern that works (regarding pattern recognition) is the one the testers think is right, and so forth, what you have is a number which is fundamentally culturally constructed and fails to capture real genius.

"Race is problematic in a number of levels. Yes it is a social construct with no real genetic basis, at least as we have constructed it."

Sorry, what? Are you talking about how things are on Earth or some other planet? Perhaps a cursory look at http://en.wikipedia.org/wiki/Race_and_genetics might help.

So if race is fully based on genetics would it be equally correct to say "Barack Obama is white" as it is to say "Barack Obama is black?"

Why or why not? And which matches our social constructs?

> So if race is fully based on genetics would it be equally correct to say "Barack Obama is white" as it is to say "Barack Obama is black?"

Yes, of course. He's a mixed race. I don't know why people call him the first black president - he's exactly half and half, with equal claim to either race.

Unless you subscribe to http://en.wikipedia.org/wiki/One_drop_rule

> Yes, of course. He's a mixed race. I don't know why people call him the first black president - he's exactly half and half, with equal claim to either race.

You have just shown that social construction of race is not the same as your biological base. If most of the population agrees "he is black and not white" and biology says otherwise, then the social construct is unmoored from biology.

I don't think one has to go as far as a one drop rule, though that was a construction that has existed in some parts of the country at some times (as have other tests, like the brown paper bag test). Things are more nuanced today and it would be more accurate to say he is black because:

1. He claims to be black, and

2. Society recognizes his claim as legitimate.

Since these are organic social constructs, they don't need hard, strict, articulable rules any more than English grammar does (I suppose someone, probably me (note the case shift contrary to formal grammars), will have to point the reader to actual articles by linguists on this topic).

I decided to respond to this part of the thread to keep discussion in one place.

---

So your point is that since 2.9% of US population are not single-race (estimate of mixed race in the US [1]) the other 97% aren't either?

By the same token you could also say that there are no heterosexuals in the world since roughly 3% of people are gay.

I remember some Hispanic comedian who said in the 90s - "In the future everybody will look Filipino". He had a point - if humans spend a few thousand years mixing it up, the end result will look Brazilian of Filipino. But for now the majority of Earth population is of a specific definable race (which have many differences deeper than skin tone).

[1] http://economix.blogs.nytimes.com/2012/09/27/mixed-race-amer...

> So your point is that since 2.9% of US population are not single-race (estimate of mixed race in the US [1]) the other 97% aren't either?

From a biological perspective, I am not sure you can biologically define "single race" vs "mixed race" to the point where such a categorization makes sense.

I think what you can do is specify identifiable populations which share some identifiable genetic markers (say, Chinese vs Mongolians, or Scandinavians vs Mediterraneans).

So what I am saying is that 2.9% of Americans are socially of "mixed race" but population groups are much more porous than that, and that does not match the biological realities.

This doesn't mean one cannot come up with a biological definition of race, but I don't think it will match our social definitions very closely.

"I think what you can do is specify identifiable populations which share some identifiable genetic markers (say, Chinese vs Mongolians, or Scandinavians vs Mediterraneans)."

Yes, and likewise you can further extend it - grouping Mongolians & Chinese into one group and various sub-Sararan tribes into another. And voilà - you've got yourself nice Asian and African races. Because they will still share these distinct identifiable genetic markers.

"This doesn't mean one cannot come up with a biological definition of race, but I don't think it will match our social definitions very closely."

Our social definitions are pretty spot on, if you don't try to come up with exceptions (ie Barack Obama et al). Jay Z is black and Seinfeld is white - how complicated is that?

Me: Blue and Red are different colors. There is a branch of physics called optics that deals with it.

You: But what about Brown - is it blue or red? I guess you can't answer, therefore colors don't exist and are instead social constructs.

But the point is:

Society unambiguously recognizes Obama as black. Biology does not. Therefore if you say society is wrong, you are just saying the social construct of race is wrong, and reinforcing the idea that race in our society is socially constructed and independent of any biological idea you can put forward.

The larger question would be for shades that we might accept are "red" but don't necessarily meet the wavelength definition. Additionally human populations are not like wavelengths in that they are dynamically defined. You can find all sorts of populations with identifiable genetic similarities. However, we construct race differently than that. Arguing that society is biologically incorrect in a definition of race is just saying that the social construct is unmoored from the biology.

Google: Dunning-Kruger Effect

I always find it bizarre the consistent need for a certain types of people to believe IQ tests are not real.

It's pretty much conspiracy sort of stuff.

A. It's quite a stretch that a whole scientific community can't create a test for IQ and get past the tired issues brought up 50+ years ago.

B. The concept brains unlike heights, hair, body shape etc etc etc are all the same is pretty ludicrous.

There is lots of stuff on the Flynn effect, I'm sure people can google it for themselves. But at the end of the day why is it so strange a test given to people in the 1960s scores differently for people in the 2010s. It is such a strawperson argument.

IQ's about relativity and ability in the current environment.

Scientists definitely do have a test for the so-called IQ, and it's called the IQ test – but 'Intelligence Quotient' and intelligence are not synonymous. One need only read Mensa magazine to realise that IQ does not constitute intelligence in a meaningful sense.

It's clear that brains vary, of course, but 'intelligence' is too broad a term to say, quantitatively, to what degree a person possesses it. The term encompasses:

- The ability to acquire new knowledge quickly, or to learn difficult things at all.

- The ability to, given time, make nontrivial deductions from given data.

- The ability to make rapid deductions in short periods of time about given data.

- The ability to generate multiple unrelated solutions to a problem.

- A high level of verbal proficiency.

We mean all these things to different degrees at different times when we use the term; and although these are not orthogonal concepts, they clearly do not exist in a one-dimensional space.

Certainly if IQ is taken to mean 'the raw, unchanging potential of a mind', the Flynn effect is inexplicable given that the timescales involved are far too small for our neurobiology to have adapted evolutionarily. Given that it means very little other than 'the ability to pass IQ tests', the Flynn effect is perhaps not surprising at all.

Yet more overreach in the conclusions people take away from studies such as this. So there is an environmental factor in what IQ tests measure; this should not be a surprise to anyone. The study suggests that one's environment growing up affects your abstract reasoning ability. This in no way suggests that it is completely environmental. Height is largely genetic, we all know that. Yet height is also strongly affected by environment. In fact, if you don't feed a kid, they don't grow at all. This does not mean that height has 100% correlation with nutrition and thus zero correlation with genes! People need to be careful with drawing convenient conclusions from a small amount of data.

Just to clarify; the study doesn't say that intelligence is completely environmental, but the reporter draws the conclusion that IQ test results are 'mostly environment and experience', which I'll agree is a reach.

But I'd welcome the study itself, further distancing us from the consensus, among what could well be the majority, that intelligence is innate and that IQ tests measure this quality.

Why is this a problem even? Why can't IQ be the result of both genes and the right environment and upbringing? Why is it any less valuable as a measure if it's affected by environment?

Sure, there's more to "that person is clever" than just your raw computing power, but it's a pretty damn big factor.

Here's the weasl-ey straw man (sort of):

"..some thinkers cling to the idea that IQ measures an inborn intelligence that transcends culture and schooling. If that's true, one would expect that the most abstract, "culture free" elements of IQ testing wouldn't be subject to."

The evidence supports common sense on the idea that environmental factors influence IQ (and intelligence) substantially. I doubt many "thinkers" disagree.

The evidence also support that IQ test are really hard to do correctly.

Some IQ test are better done by urban vs rural children and vice versa and it comes to what knowledge does IQ test verify and what method does it use. It's not as easy as saying its environment. Test makers suck at thinking outside their head and disabling their own knowledge expectations and biases.

I would question the premise actually that intelligence can be represented adequately by a single scalar quantity. a much better approach would be an n-dimensional vector.

The problem you get with a scalar measurement is that you end up with weighing one component of intelligence against another. IQ tests don't spend all questions measuring the same exact skills of course, so you have to have a way of weighing one skill against another. That weight itself can only be culturally derived, and therefore IQ tests in their current form have no hope in measuring innate abilities on an apples to apples basis because in the end we are comparing apples to melons.[1]

[1] Hats off to you if you got the pun, which will only be found through careful study, knowledge, and reading slowly.

You mean the weighting apples vs melons? If so, the box was a dead giveaway.

The pun is two-fold and juxtaposes Modern English and Classical Greek, which is part of what makes it relevant to my point ;-)

To explain, "melon" means "Apple" in Greek.

So you can either compare two different languages, or two different fruit but you can't do both at the same time ;-)

Sadly I was never good with Ancient Greek or latin for that matter.

Well, it still measures relative intelligence for people with similar education, and in our globalized society our educations and exposure to information get more and more similar, so it's very good at what it's suppose to do and my guess would be that is getting even better as "cultural homogenization" in the global village increases... Now, we all know that intelligence is just a variable, it does not singly predict suitability for a certain job and it's very bad at predicting things like "success in life", but it is a useful variable!

It is refreshing to see this kind of research promoted, particularly by an 'extremely learned person'. The submission reads to me as a sort of call to arms, as someone who feels that education has been a great privilege and bonus to their life. A privilege which is now unfortunately being sold off to the highest bidder in most countries.

From my own experience, learning second/third languages has had a noticeable broadening effect on thinking processes. Not in the proud sense of being 'broad minded', but noticeably better able to engage in abstract concepts.

Basically the ability to compare seemingly unrelated notions felt tangibly improved after learning how my speech in English was actually constructed, and how my words and phrases were just $variables for actual concepts. And I'm quite convinced that given the right chances in life, anybody can learn any of a thousand subjects that would have a similar effect.

The end benefit to society? Impossible to say for sure, but you can bet good money on the additional creativity being an improvement on the previous generation in terms of progress.

This article, and others like it, are great to read but one aspect I would like to see further developed would be the effect of role models and the origin of aspiration in a person. Is it even possible to give more people the desire to want to learn more? Are some people just naturally more 'competitive'? What role does gender play?

On a final note, Radio 4 this week interviewed a panel on women going into Physics, in particular. It was a ridiculously low amount, and one of the female panellists excused this with the reasoning that 'girls often prefer biological sciences'. These kind of assumptions, and a complete ignorance of any potential underlying social issues, are what get my goat. So many thanks, Colin, for submitting this.

Why do people breed dogs for all kinds of absurd characteristics, but nobody has tried to breed dogs for intelligence?

I've seen reports that some dogs achieve a vocabulary of 400+ words. What would happen if one selected for that? Could you conceivably get a dog you could have a conversation with?

If you ever spend much time in a foreign language learning a very different language (bonus points for it being non-Indo-European), you will note that there is a huge gap between understanding hundreds or thousands of words and being able to carry on a conversation.

Of course. But it's a good starting point.

I think that one of the huge problems is that there has never really been to my knowledge a systematic look at the epistemological limits of IQ tests. However just looking at it quickly I can spot two huge epistemological issues which make IQ tests more or less worthless for objective study, and more or less cultural artifacts.

The first, which I have mentioned in a few posts is that at some point you end up comparing apples and melons (the pun for those of you who didn't get it is that 'melon' is Classical Greek for 'apple'). These things may be things where we can make connections but any cultural weight we place on one relative to another is just that, a cultural weight. For this reason intelligence, even if IQ tests measure what we expect them to (and I am not sure of that, see below), the results cannot be objectively reduced to a single number. You would really need to look at intelligence as a multi-dimensional vector with the understanding that we might be missing some aspects anyway.

The second problem though that I can spot immediately is that some skills they want to test cannot be objectively tested. The big one that comes to mind is pattern recognition. If you haven't read it, I think Heisenberg's "Physics and Philosophy" is a must read here because his discussions of the relationship between data and theory are directly relevant to this.

When we recognize a pattern what we actually do is we project assumptions onto the data and find a pattern which fits both the assumptions and the data. This is the process Heisenberg articulates regarding the formation of scientific theories. The pattern recognized is both a product of our assumptions and the data and therefore patterns are at once recognized and created. Of course Heisenberg was talking about his forte (theoretical physics) but the same applies here because this is a human process.

The simple fact is that several competing sets of patterns can be made to fit the data. For example, take the following set of items:

Hammer, log, ax, saw

Which one doesn't belong?

One pattern might be to that "hammer, ax, and saw" are the pattern and log is the outsider. Another pattern might be "I can build something with the hammer, the log, and the saw, much more than I can with any other combination so the ax doesn't belong." Therefore such a question has at least two legitimate answers because two different set of assumptions can be projected onto the set with very different results. Now typically we are going to say one of those (the first, "log" doesn't fit) is correct and the other is wrong, but that doesn't actually test the ability to recognize patterns so much as it tests the ability of the test taker to figure out the answer that the author wants. This makes pattern recognition fundamentally culturally determined and if you can't get away from this you can't quantify this through standardized testing.

But let's take this further. We typically refer to a genius as someone who finds useful patterns that nobody else does. The IQ test then, if we accept Heisenberg's epistemology, should actively select against genius in these areas-- either the patterns must be trivial, or they must be sufficiently complex you cannot be sure whether a "wrong" answer in fact does not have a recognizable pattern to it. So you end up having to throw out a good chunk of the test here as being very closely culturally determined (and basically an exercise of mind-reading the author).

The first problem could be solved by expressing intelligence as a multidimensional vector, but the second is much more problematic. How do you know that the test answer is right and the test taker is wrong when it comes to two competing patterns seen in the question?

IQ tests are pretty sophisticated and, as the theory of intelligence or more correctly intelligences has evolved, the tests have evolved alongside.

Each test is really a suite of dozens of subtests, each of which tries to tease out some aspect of mental function. Some of it is pattern matching. Some of it is memorise-and-recall. Some of it is tests of "general knowledge". And on and on. Think of what you consider to be a distinct kind of mental performance and somebody, somewhere, has built a test for it.

I think it's important to distinguish between three things in discussions of IQ:

1. Are some people "more" intelligent than others?

Taking the fuzziest, least precise view of what "more intelligent" or "less intelligent" means, I think yes. It would be very strange that the architecture of the brain and environmental factors would lead to perfectly equal outcomes in all subjects. And it doesn't seem to fit what we can casually observe around us.

2. What is intelligence, and what are the components of intelligence?

A very large body of research discusses, explores, argues and delves into these simple questions. In general a single "IQ" figure is merely a statistical index of the subtests. The sub-category indices are much more revealing.

3. Can the tests actually accurately sample these components?

Also hotly contested. The closest we can come is to compare the test scores statistically to other outcomes, or to other tests. IQ tests which correlate highly with life achievement (factoring out stuff like socioeconomic background, sex, race and mental health) are preferred over those that don't. You can argue that it's circular, and you'd be right.

Welcome to psychometrics, where there is no outside world available for objective instrumentation.

> Taking the fuzziest, least precise view of what "more intelligent" or "less intelligent" means, I think yes.

No argument there.

> In general a single "IQ" figure is merely a statistical index of the subtests.

Which is why I point out that any weighing between subtests are relatively problematic in generating that number.

> The sub-category indices are much more revealing.

100% agreed there.

> The closest we can come is to compare the test scores statistically to other outcomes, or to other tests.

But the problem is, the ability to accurately sample these components may not be the same for each component. For example, if you want to measure coordination of multiple pieces of information short term memory, the problems are very different than if looking at pattern recognition. So I would caution against a single answer. However if we don't understand which subtests are particularly problematic and what the epistemological problems look for each subtest, then many of the really interesting questions about these tests (for example why do Asians score higher than whites) are forever out of reach.

> Welcome to psychometrics, where there is no outside world available for objective instrumentation.

Yep.

This is not evidence for anything - the theory kind of checks out but it can be argued that there are other confound variables and that the direction of causality is not as simple as they make it out to be. Also just because IQ can be (slightly) trained it does not mean that it does not measure 'intelligence' - what is the problem with intelligence being trainable (not that IQ is perfect or anything)? And lastly - yay for meta.

Does anyone here really believe that IQ tests have any value?

http://en.wikipedia.org/wiki/Intelligence_quotient#Social_ou...

I do. It is correlated with all kinds of positive outcomes in life. Given a somewhat similar cultural upbringing, there is clearly something different about someone with a 140 IQ vs someone with a 100 IQ. This fashionable meme that IQ is meaningless has no basis in fact. It certainly doesn't measure innate potential exactly, but then not too many people would say that these days anyways. However, it does correlate with something regarding one's mental capacity, and this correlation also correlates with other important outcomes.

It is correlated with all kinds of positive outcomes in life.

I beg to differ, and I have this quote to rationalize it:

"Pain and suffering are always inevitable for a large intelligence and a deep heart. The really great men must, I think, have great sadness on earth." -- Fyodor Dostoevsky

Would I rather be stupid and happy? No! And maybe I'm not of "large intelligence" in the first place, but I'm intelligent enough to often feel very disconnected and alienated from the society around me. I learned how to tell people what they want to hear, but inside I often feel like I'm trapped in some kind of zoo. The heat death of the universe is the only grown up thing I actually get to deal with, and I get to deal with it alone, while most everybody else is either drunk on brain chemistry or not playing in that league. I feel like I'm sitting in some kind of waiting room where there is nothing to do. If it wasn't for things like music, dancing, cats, dogs and toddlers, there is no way I would have made it this far. Flowers and sparrows nourish me, people are a chore. Capitalism? Don't even get me started, unless you're willing to give me an army with which to solve the issue.

And now I'm going to completely contradict myself and say that intelligence is like the horse power of an engine; yeah, it matters, you cannot go anywhere without an engine, and you can't go fast or far with a crappy one. BUT what might matter even more is the steering wheel, and I totally suck at using that. I speed around a lot and end up in ditches, and I think there is something to the "saying" that more intelligence makes bigger mistakes possible; I am a living example of that.

I know I am just whining. My personal inability to enjoy life and other people doesn't have to mean intelligence isn't generally positive; but still, when it comes to society and living and working together, it's all rather relative, and being too smart or too stupid both can easily lead to inefficiencies and heartache. Humanity as a whole might have better outcomes if it was more intelligent as a whole, but individuals being too far ahead cannot do much other than acknowledge all the beautiful things that could and should be but never will.

That said, I would never ever want to swap. I'm sure less intelligent people have things they are sad about as well, while being worse equipped to reflect on them, and I do not envy that at all.

I've never liked the idea of IQ tests, but they do correlate quite well with future achievement like jobs, income, incarceration rate etc. That is why they are still used:

http://www.iq-tests.eu/iq-test-Practical-validity-800.html

I'm in the UK and I've never encountered IQ tests as part of education or any work environment.

[NB For my own amusement I did sit a test as a teenager and got a stupidly high score, which I can actually explain pretty easily, that rather demonstrated how shallow the tests seemed to me]

I've also never directly encountered anyone discussing their IQ - which would strike me as very silly indeed.

It's probably illegal to give IQ tests to potential employees. It is illegal in the States.

Under what theory? Disparate racial impact? Before I get downvotes for that it is worth noting that racial groups do have remarkably different IQ test results in the US. Whether that says anything about intelligence or genetics is hotly debated in some circles. It isn't clear how much is cultural, whether there are subtle cognitive differences between racial groups, and even of so whether this corresponds to any intelligence difference.

In the US, IQ tests are (in some states) used in public schools during part of a process to identify students who have unique potential that may be realized if special consideration, privileges, and opportunities are given to them. So called "gifted" programs.

For example, I used one of these programs to spend half my highschool career playing Doom on school computers, and spend the other half of my highschool career taking courses at the local branch of the state university.... Yeah, not a perfect system.

Using IQ tests during hiring is legally risky in the US because it puts you at a high risk of discrimination lawsuits. Subsequently, where you see them in the US, they are masquerading as something else (think "puzzle"/"outside the box thinking" interview questions for programmers).

IQ correlates nicely with the amount of lead child was exposed to while growing up and the probability such grown up child will be involved in some kind of petty crime activity later on.

Such measure must have some value.

People who had learned maps to become proper London cabbies tend to be better at memory tests later on. So should we conclude that culture of memorizing stuff they experienced is responsible for their high marks? And somehow that biology determining their physical aptitude isn't involved?

This does not follow. Their brains changed visibly on MRIs over the course of map learning.

Of course that's the culture that makes people more intelligent. But you can't dismiss this gain of IQ as not real or non biological. That's probably the only real IQ gain that we ever got since our animal ancestors decided that it's better to live in groups, and they themselves became their smartest competition. Culture is just a fancy name of all of human interactions.

I always considered the intelligence as an abstract concept. Like a way to named the group formed by curiosity, flexibility, adaptation and others capacities to understand relations and interactions between elements.

By the abstract nature of that set of attributes, I don't understand how it could be measured.

By the way, what is your definition of intelligence ?

This actually confirms that IQ measures intelligence, just not merely innate intelligence. Which we already knew.

Well the article claims IQ measures the capacity for abstraction, which isn't necessarily the same as intelligence: the best programmer isn't the one who always starts with a UML diagram.

Huh? Programming is basically abstraction taken to the extreme. It's writing repetitive code and then going "you know what would make more sense? If I could abstract this repetitive part into another construct and then reuse it". A programmer who has bad abstraction skills will just copy paste a few times and modify the necessary parts. And forget to change stuff in a few places probably.

I don't think IQ tests are immune to grade inflation (http://en.wikipedia.org/wiki/Grade_inflation), so I am inclined to think of Flynn effect as a special case of it.

The only definition of "intelligence" that matters is "whether you seem intelligent to other people". That's what an IQ test attempts to measure, not the raw processing power of your brain

I'm glad that there's further evidence that IQ doesn't correlate with intelligence. I've never been a fan of quantifying intelligence or other characteristic attributes.

Reply on news.ycombinator.com