Item — HN Reader

Malcolm Gladwell states in Outliers that there's really no correlation between high IQ and performance. After a certain IQ, I think it's 120, the rest is about emotional maturity and how deeply your understanding goes.

I think that in the future we'll come to realize that IQ is actually a silly metric and that this "understanding" is where it's at. Point of view (understanding) doesn't give you more intelligence (Alan Kay's proverbial 80 IQ points), but instead intelligence can help you to gain understanding.

9 comments

For some reason I want to read heavily between the lines of the wikipedia summary, and I find it incredibly sad. I mean in the first place my knee jerk reaction is to equivocate "I'm still good enough, even if I didn't solve diffeq problems at 6. My son's okay even if he still doesn't read at the age of 4."

But I don't want to imagine the kind of expectation people put on this kid (when he was a kid). Cynically, I suppose as I read I thought "oh yeah what's he done now? We still have cancer and no unified field theory. Does P=NP?! huh?" and yet under that was the expectation "well this guy should be curing cancer and resolving open problems!"

No wonder he escaped to something ordinary (poor Civil Engineers, though, what a bad rap that is ... my profession is the most ordinary!) People haven't changed much since the dark ages (or before). They still expect miracles and magic to cure their woes, they just expect it now from Science.

It agrees with me that accomplishment comes after the 120 IQ...heck I'm barely above the marker.. shwoo. But it comes in increments. It comes in mostly in pieces we have to put together. Einstein didn't make the world a better place single-handedly, it was the engineers who came after and created GPS (for instance) that did it. There's no reason this guy should have created miracles with his IQ, we probably don't have enough of the base research to solve that (whichever) problem, yet.

Instead, he's improved the world a bit (quite a bit in the sector of civil engineering apparently). Honestly, the universe is miracle enough. The fact that we can make it slightly better by our actions and increase joy in the miracle of existence at all is a good thing. Some people do a better job of that with much smaller IQs.

"On November 2, 1967, at age 4, he solved an advanced stochastic differential equation."

... and then gets a PhD in physics at age 16. Although the IQ score might be outdated/irrelevant - there is something special about this man and if we don't strive to quantify it we won't be able to reliably reproduce it.

I'm somewhat confused. What does being able to label something have to do with reproducing it?

I may be able to label the heart in a person, or perhaps distinguish an exceptional sonata from something that isn't, but that doesn't mean that I can create a hart, or play that song.

Although you could see it as reproducing this human (breeding, eugenics) you could also look at it as discovery and training. Sort of like being able to spot a really great runner at age 5 and getting them to the Olympics by age 18, etc.

I think if we can understand (since it probably isn't IQ) the "real" reason this person has such above-average abilities then we can find ways to test for it and help other people with these abilities develop fuller potential, and hopefully humanity gets something back - in both the arts and sciences.

I sent that quote to everyone in my stochastic calculus class. I couldn't be the only to feel that dumb today!

The original purpose of IQ tests was not to predict who would go on to the most brilliant career, but to help teachers recognize who would need special help in classes. In that context I think the distinction between 120, 150, and 180 IQ is meaningful.

The distinction between 150 and 180 IQ is meaningful? Really? If such a difference is reliably measured, and not easily trained by either person (e.g. by devouring a few books on vocabulary and brain puzzlers in the week before the test), then what is the meaningful difference? Both people are going to be bored out of their mind at school and are going to have to be basically self-reliant for their actual learning process for much of their lives. Feynman supposed had an IQ in the low end of that range, and a given person at the higher end probably isn't as smart in any real sense as Feynman was. Whatever the difference is there, it isn't IQ.

Feynman had the advantage of built-in syntax highlighting, a gift I would absolutely love to have:

"When I see equations, I see the letters in colors - I don't know why. As I'm talking, I see vague pictures of Bessel functions from Jahnke and Emde's book, with light-tan j's, slightly violet-bluish n's, and dark brown x's flying around. And I wonder what the hell it must look like to the students."

http://en.wikipedia.org/wiki/Grapheme%E2%80%93color_synesthe...

Feynman had an IQ of 127.

He wouldn't have gotten into the gifted class I went to as a kid.

He was, however, free to think, and act on those thoughts, and these he practiced into an art. That is more marvellous to me than a number on a test.

I don't know if the distinction between a tested 150 and 180 IQ is meaningful, but that between those levels of intelligence is. I'm in the 150 range; I've met some who are probably in the 180 range (IMO gold medalists) and they are noticeably smarter than I am.

That said, even a 200 IQ won't overcome a lackluster work ethic, mental illness, or social adversity (internal or external).

I'm glad I don't know my IQ, I can keep imagining I'm in the 500 range ;p

IT'S OVER 9000

The distinction between 150 (barely recognizable as a validated score on current tests) and 180 (NOT a validated score on any test, ever) is less meaningful than many suppose:

"Put into the context of the psychometric movement as a whole, it is clear that positive extreme of the IQ distribution is not as different from other IQ levels as might have been expected. . . . While 180 IQ suggests the ability to do academic work with relative ease, it does not signify a qualitatively different organization of mind. It also does not suggest the presence of ‘genius’ in its common-sense meaning, i.e. transcendent achievement in some field. For these kinds of phenomena, IQ seems at best a crude predictor. For anything more, we will have to look to traditions other than the psychometric and to variables other than IQ."

Feldman, David (1984). A Follow-up of Subjects Scoring above 180 IQ in Terman's Genetic Studies of Genius. Exceptional Children, 50, 6, 518-523.

IQ scores, especially in childhood, vary over the course of a test-taker’s life, sometimes varying radically. Deviation IQ scoring was originally developed to make for more stability of scores over the course of childhood. Nonetheless, deviation IQs for children can also change considerably over the course of childhood (Pinneau 1961; Truch 1993, page 78; Howe 1998; Deary 2000, table 1.3). "Correlation studies of test scores provide actuarial data, applicable to group predictions. . . . Studies of individuals, on the other hand, may reveal large upward or downward shifts in test scores." (Anastasi & Urbina 1997 p. 326).

For example, young people in the famous Lewis Terman longitudinal Genetic Studies of Genius (initial n=1,444 with n=643 in main study group) when tested at high school age (n=503) were found to have dropped 9 IQ points on average in Stanford-Binet IQ. More than two dozen children dropped by 15 IQ points and six by 25 points or more. Parents of those children reported no changes in their children or even that their children were getting brighter (Shurkin 1992, pp. 89-90). Terman observed a similar drop in IQ scores in his study group upon adult IQ testing (Shurkin 1992, pp. 147-150). Samuel R. Pinneau conducted a thorough review of the Berkeley Growth Study (1928-1946; initial n=61, n after eighteen years =40). Alice Moriarty was a Ph.D. researcher at the Menninger Foundation and describes in her book (1966) a number of case studies of longitudinal observations of children's IQ. She observed several subjects whose childhood IQ varied markedly over the course of childhood, and develops hypotheses about why those IQ changes occurred. Anastasi and Urbina (1997, p. 328) point out that childhood IQ scores are poorest at predicting subsequent IQ scores when taken at preschool age. Change in scores over the course of childhood shows that there are powerful environmental effects on IQ (Anastasi & Urbina 1997, p. 327) or perhaps that IQ scores in childhood are not reliable estimates of a child’s scholastic ability.

REFERENCES

Anastasi, Anne & Urbina, Susana (1997). Psychological Testing. Upper Saddle River, NJ: Prentice Hall.

Deary, Ian J. (2000) Looking Down on Human Intelligence: From Psychometrics to the Brain. Oxford: Oxford University Press.

Howe, Michael J. A. (1998). Can IQ Change?. The Psychologist, February 1998 pages 69-72.

Moriarty, Alice E. (1966). Constancy and IQ Change: A Clinical View of Relationships between Tested IQ and Personality. Springfield, IL: Charles C. Thomas.

Pinneau, Samuel R. (1961). Changes in Intelligence Quotient Infancy to Maturity: New Insights from the Berkeley Growth Study with Implications for the Stanford-Binet Scales and Applications to Professional Practice. Boston: Houghton Mifflin.

Shurkin, Joel N. (1992). Terman's Kids: The Groundbreaking Study of How the Gifted Grow Up. Boston: Little, Brown.

Truch, Steve (1993). The WISC-III(R) Companion: A Guide to Interpretation and Educational Intervention. Austin, TX: Pro-Ed.

It seems like you know a whole lot about this. I once heard a Math professor say that children were able to use more methods for doing basic arithmetic when they first started school than when they'd gone to school for a year, or something of that nature.

I've also been through public school, and I know that the way they teach stuff is not modern in any way. It is almost the same school system that Prussia had, that later was adopted by countires like the USA. A school of compulsion, instituted to teach students the "virtue of obedience" and to make them parts of a bigger machine (like the war machine). So the teaching method of choice naturally became to have the students do simple tasks obediently at their desks. This can be seen in modern American classrooms aswell (yes, I went to one for a year). Think back to American History, what did you do really? Did you copy down, word for word what the teacher said? Was those chores teaching you anything on how to think when it comes to reviewing history? No, it couldn't have, you were just fed words that you didn't even have time to analyze. How could that have taught you anything?

I think the same goes for Math. In Sweden you didn't learn the principles of Math in the first ten years (this is important, since Montessori Schools (atleast those who are run by sane people) are able to teach kids the actual concepts behind Math at a very young age). You had 100 similar problems to solve, and you had the solution on a paper, you just had to do that a 100 times or more and remember long enough to pass the upcoming test, then you could forget about it. This is BAD for problem solving, and I can only thank programming for not stripping of the pleasures of REAL problem solving. So naturally, when a child goes to do an IQ test, which is basically a test in problem solving, he/she wouldn't know how to do it, because he/she would have been used to not having to solve problems.

I think this would explain a lot, I think that IQ can be used to explain a lot of things if the world actually prepared you for them. Am I wrong or do I have a point? Anyway, that's my theory.

I once heard a Math professor say that children were able to use more methods for doing basic arithmetic when they first started school than when they'd gone to school for a year, or something of that nature.

My anecdotal experience correlates with this. First grade, with its required use of "bubble math," was very detrimental to my arithmetic abilities I'd acquired at home and during Montessori-style preschool.

So naturally, when a child goes to do an IQ test, which is basically a test in problem solving, he/she wouldn't know how to do it, because he/she would have been used to not having to solve problems.

Some IQ tests seem to test certain kinds of problem solving, like identifying patterns in a grid. Some few individuals may be able to identify the patterns without any prior experience, but I suspect most who answer such questions correctly do so from knowledge of a general class of problems, not intuition.

Purely subjectively, I feel the busy work in school made me stupider, whereas once I've been out and learning on my own, my mind feels better and sharper, and I'm more interested in learning.

My opinion is that two really big factors are desire and free will. People need desire in order to really excel, and they need to think for themselves to quickly see good solutions. Both of those are hard to capture on an iq test.

to help teachers recognize who would need special help in classes

As opposed to SATs, which were designed to determine who was most likely to succeed in post-secondary education (which again has relatively modest correlation with career success).

One of my psychology professors pointed out that income and IQ have a positive, but far from perfect correlation. He said we all could probably think of people who aren't the sharpest who've made great money. He also pointed out that those with high IQs often seek careers they think will be intellectually challenging, but not the most financially rewarding ... like becoming professors.

This also goes to show how much determination matters for a startup. Yes, being bright matters, but being determined matters more.

An uglier aspect of it is this: although people with 150+ IQs, contrary to the nerd stereotype, usually get along fine socially, they'll virtually never be selected as leaders, and always be outsiders. People usually aren't comfortable with a leader who is that smart.

We evolved in small tribal groups of 20 individuals per generation, so we look to the 90th-95th percentile (6'1, 120 IQ) for leadership, but extremes and "giants" frighten us. 99.5+ percentile is a threat-- possibly some other hominid that is going to eat us.

Bill Clinton is supposedly scary-smart, at least in terms of being quick-witted and able to multitask very effectively. It wouldn't surprise me at all if his IQ were over 150.

Edit: see http://www.time.com/time/magazine/article/0,9171,987883,00.h...

"It's an almost scary mind, that of a multitasking wizard who plays hearts while he talks on the phone with a head of state, who sits through a dense briefing on chemical weapons intently doing a crossword puzzle, only to take reporters' questions hours later and repeat whole sections of the briefing word for word."

See also this page from Robert Novak's autobiography: http://books.google.com/books?id=7o5Cpjkk4qsC&pg=PA547&#...

Sure. The same goes for Barack Obama. And this is why they're facing one hell of a revolt from the morons, who will not accept such people as them for leaders.

There are legitimate reasons to disagree with both Clinton and Obama, but the real problem they face is that they make the dumbs feel bad about themselves. Said dumbs elect manipulative bastards who play to their complex, and said bastards launch ad hominem attacks and make it almost impossible for them to get any work done. This is what the Clinton sex witch-hunt and the modern-day Republican make-Obama-fail-at-all-costs efforts are really about: stickin' it too those smarty-pants lib'ruls.

It must be comfortable to think that those who disagree with you does so because they are intimidated by awesomely intelligent liberals such as your heroes (and perhaps yourself?). Has it ever occurred to you that they might feel the same way about you?

It must be comfortable to think that those who disagree with you does so because they are intimidated by awesomely intelligent liberals such as your heroes (and perhaps yourself?).

I don't think that, but I think anti-intellectualism is a major reason why Clinton and Obama both got mired in ad hominem attacks that made it difficult to get anything done.

That's different than claiming that people with IQs of 150+ "[will] virtually never be selected as leaders, and always be outsiders. People usually aren't comfortable with a leader who is that smart." Regardless of his accomplishments or the obstructions thrown up by Republicans, a _lot_ of people were comfortable having him as a leader.

Does IQ correlate with leadership ability?

Intelligence dedicated to solving abstract or impersonal problems is worthless for a leader. Therefore, the kind of intellectual talents that show up on an IQ test are not very useful. Someone who takes a problem-solving approach to the world will not make a good leader. Someone who takes a problem-solving approach to his parents, friends, classmates, and so forth will do well.

That's because leadership consists in large part of stating and forging a group consensus, perhaps not in that order. A leader rarely suggests a course of action that doesn't immediately make sense to other people. If education and elucidation is required, he does that beforehand. A leader doesn't bother with courses of action that his followers aren't already one little nudge away from accepting, and then he provides that nudge. For this reason, it's best for a leader not to devote too much intelligence to actually solving problems -- he will just come up with useless and distracting ideas. A leader with a lot of credibility can get people to accept a new course of action solely on his own credibility, but if it doesn't work out, he loses a lot of credibility. He mostly needs to stick to courses of action that are credited to other people or which just feel right to his followers.

Leading by surpassing excellence is an invention of propaganda. Only after you already have a lot of power and loyalty can you start manufacturing such a myth.

That's a leader in the popular sense. But if being such a leader means he can't think deeply, then he'll have to rely on the intellectuals to provide him with the deep ideas he needs to lead.

At that point, who's the leader?

Fair point. The creative visionaries are the people who should be leaders in an ideal or even morally acceptable society, but almost never the people who have the capability to grab power.

no left handedness does.

It depends how "leadership" is defined.

By and large, I am a strong believer in rule by the smart, although culture is a lot more important than IQ. Probably 20-30% of those in the ruling class have a high enough IQ (125+) to be there but less than 1% have sufficient culture. Paul Wolfowitz and Dick Cheney come to mind as examples. They're so uncultured they probably eat steak with their fingers; that such louts could rise to such high levels of influence is a serious problem. And our society is such an ugly mess because people like them (uncultured people, with ugly minds) rise to power and remake the world in their image.

I think it's also obvious that most business executives are selected for people-pleasing and crude optimizations (next-quarter profits) rather than vision, creativity, and humanity.

The traditional two-tier managerial structure (although excruciatingly flawed and less-than-useless in practice) had the right idea. At one degree above the average (110-125) you have the managers, whom the line workers look to for leadership, but who do not alienate them with their intelligence. At two degrees up (125-145) you have the executives, who can lead and mentor the managers, and whose closer connection to "the metal" gives them a strong voice in the decision-making. At 3 and more degrees, you have strategists and academics who present the information and advise the decisions. However, modern corporate society doesn't work this way; the executives are often mediocrities, the "advisor" class isn't really listened-to, and most of what goes on in academia is self-referential bullshit.

"Rule by the smart" is not flawless, of course. Bigger problems than a lack of intelligence in our ruling class are the lack of integrity, vision, and compassion... which are also much harder to measure-- although all of these correlate are correlated much more strongly with IQ than with "leadership" as measured by our current selection process. Therefore, an "IQ meritocracy" would be an improvement. Between a set of 140+ IQers (among whom there would be some who are unfit to lead) and the people running the show now, I'd bet on the first class without hesitation.

After a certain IQ, I think it's 120, the rest is about emotional maturity and how deeply your understanding goes.

It would be a mistake to underestimate the importance of people like this. He may not be the largest "producer" or whatever your personal metric for being successful is, but even solving a small problem that you or I won't ever be able to will have a unique, lasting, permanent effect on the world.

I tend to agree. The fact that your IQ metric can be swayed by lack of sleep, bad experience in the day of test, etc. make it an inaccurate way to measure intelligence precisely.

"... there's really no correlation between high IQ and performance. After a certain IQ, I think it's 120, the rest is about emotional maturity and how deeply your understanding goes."

Carl Friedrich Gauss. General life "performance" at popularity and piling up money may (may) top out at 120, but there are some jobs where performance means being able to see through a brick wall of symbols like it is a window pane.

Any person with an IQ of 120 can learn and reproduce Euclid's proof of the infinitude of primes, especially because most educated people have familiarity with the most elementary number theory concepts (divisibility, primality).

Less than 1/10,000 of people with IQ > 120 could have come up with it at that time, keeping in mind that the only language one had to express numbers was geometry-- no algebra or number theory existed as we think of them today. Of course, whether this rare ingenuity shows up on an IQ test is an open question.

The same holds for something like Godel's theorem. I'm starting to understand it-- he's encoding a Lisp in number theory-- but there's no way I could have come up with it in 1931.

After a certain IQ, I think it's 120, the rest is about emotional maturity and how deeply your understanding goes.

I know you don't mean it this way, but this comes off as very smug. There are lots of reasons why most very-high-IQ people (170+) seem to underperform their potential-- some good, some bad-- but they're a lot more complex than a lack of emotional maturity and innate social acumen.

Lacking "emotional maturity" doesn't necessitate simplistic. Would you mind expounding upon what other factors there are?

First of all, I think that people with high IQs tend to have a stronger tendency toward mental illnesses and related patterns such as rumination, hypersensitivity to rejection, and low self-confidence. I think many of the underperformers have what might be considered subclinical mental illness (but is more accurately described as the normal exaggerated sensitivity of creative people and adolescents).

So, they're more "fragile" off the bat, and this is probably neurological in basis. We can provide more support as a society, but there are obvious limits regarding what can be done here.

Second, society is generally bad at recognizing talent and connecting it with the resources that could be used to actualize it. Those resources tend to go to people-pleasers and horse-traders rather than to those who can make the most use of them (who are almost always viewed with suspicion). This is true as much in academia as anywhere else.

Combine the general stupidity of society and the hypersensitivity of the sharpest intellects, and you have a lot of very smart people who get discouraged and bitter, then lose all desire to play.

Is that a belief support by anecdote, or have you seen some study supporting it?

This article is pretty good on covering the beneficial neural correlates of depression: http://www.nytimes.com/2010/02/28/magazine/28depression-t.ht... .

Whether they bear on IQ itself is an open question, but the fundamental concept is that certain mental habits that produce depression also produce focused thinking (which is a necessity for people who want to achieve at high levels).

Maybe one explanation is that people who can focus very well on a problem can see it best and solve it best, thus have high IQs. Since their minds focus so well, it is hard for them to lose focus and loosen up, essential for socializing. Thus, people treat them as weird, they feel rejected, and grow bitter.

Very likely. The intense, problem-solving focus doesn't hold well in social interaction.

A common social quirk of smart people, when they're young-- say 5-10, is to come out with a really good retort 5 seconds after it was appropriate and when it's no longer funny (because it requires a conversational "rewind"). This is because they've put a lot of energy and focus into saying the "right" (funny, most socially beneficial) thing instead of going with the flow.

I think it's cute that you believe something that Gladwell put forth to the world.

http://studentcareers.learnhub.com/lesson/6941-iq-ranges-of-...

Your link doesn't offer any sort of evidence against Gladwell's contention (which isn't original to Gladwell, and which on the basis of negligible evidence I suspect is an oversimplification but not far from the truth).

In fact, arguably it offers a little evidence against what Gladwell says. If IQ above ~120 makes no difference to performance, and if "career" and "performance" are close enough to make it worth while citing this thing at all, then you ought to see no little difference between {occupations substantially found among people with IQ ~120} and {occuptations substantially found among people with IQ ~130}. Now, look down the 120 and 130 lines of that plot, and notice that you see quite different things in the two instances. (The foregoing is a bit of a simplification, but it'll do. Anyway, I see no way to read the graph as providing evidence that IQ 120 is better than IQ 110, but IQ 130 isn't better than IQ 120, in terms of likely success/performance.)

The estimate that the IQ threshold for adult eminence is about IQ 120 appears to come from Arthur Jensen. It has been repeated by Hans Eysenck and is most frequently written about by Dean Keith Simonton, who cites a specialized study on exactly that issue in one of his latest books.

Locally, I'm in an ongoing discussion, partly in person and partly by email, with Thomas Bouchard on this point. I think the additional help of an IQ score above 120 is negligible compared to other factors that are known to be important for adult eminence.

Your graph only shows that IQ matters up to about 120. I agree that Gladwell wildly exaggerates his case, but your statistics don't contradict it at all.

Malcolm is certainly not the first person to put that idea out there, and I think there's some credence to it. I read something else recently (based on a Stanford study) that said that the greatest predictor of success wasn't IQ, but rather self discipline: http://www.newyorker.com/reporting/2009/05/18/090518fa_fact_...

Certainly someone with a high IQ AND self discipline could go farther than someone with just one or the other, but IQ by itself is not a meaningful gauge of someone's potential.