Studies correlating IQ to genius are mostly bad science
97 comments
·February 23, 2025kordlessagain
juniperus
I think when it comes to "super recognizers," their fusiform face area (FFA) in the temporal cortex of the brain may be better/differently connected. The opposite end of this spectrum is prosopagnosia, where you can't remember any face. There is also prosopometamorphosia, where you perceive faces to be wildly disproportionate and demon-like (google it). So these might not be necessarily correlated with IQ as much as they are related to specific neural pathways/circuitry. Same can be true for visualization of data. There are many autistic savants who can perform specifically remarkable feats, but this isn't really translated perfectly onto the IQ test. The IQ test is probably an imperfect measure of all types of intelligence, but for general intelligence, it provides some level of useful data, but I don't think you can say that data covers all different specific scenarios like those described, sometimes it certainly might. There are probably people who can score 150 on an IQ test, but are actually unable to do any productive creative activity meaningfully.
crooked-v
For an anecdotal case:
I've never taken an actual IQ test but I've always done very well at academic testing in the abstract with basically any subject I encountered (with some knock-on effects; for example, nobody ever teaching how or why to study anything, leading to a crash-and-burn in college).
I have almost zero visual recall, to the point that I would describe myself as mostly aphantasic. For example, without a reference photo, I'm literally unable to describe most peoples' faces outside of the most generic details ("white guy with uhhh, intact face and no beard?"), even if I know them well.
d-lisp
I always had 19.5/20 at school no matter the subject, if the subject involved logic and even if I didn't learn anything. On the other part, my grades in other subjects such as history were much more "ordinary". I skimmed through the academic part of my life without paying any attention to what the teachers were telling, drawing stuff in a notebook.
On the other hand, I cannot remember simple things such as where to go to <enter the name of a place> and I often have to wander and explore when I ride my car ...
Also, if you ask me what I did yesterday, it is highly probable that I won't be able to tell.
Also, I never know what day of the week we are, and sometimes I don't even know what month we are.
For every other aspect, learning new things is like learning things I already know.
vjk800
> For example, my neighbor in California - he had an ability to look at an equation and actually visualize the plot of it like a graphing calculator. That is certainly intelligence.
That is certainly something you develop - and even make a conscious effort to develop - if you work a lot with equations. Some level of mathematical intelligence might be a necessary prerequisite for learning it, sure, but it is something that quantitatively inclined people can learn to do.
Source: I'm a theoretical physicist, and visualizing equations is certainly the traditional way work in this field (although nowadays computer simulations are also there)
flowerthoughts
I think the way our memory works shapes a lot of what we become. I can remember numbers, algorithms, weird words and 3D shapes. But struggle with music (piano), names, what I did a week ago, and what your birthday is. That shapes what I become interested in, because I can't progress my abilities if I can't remember. E.g. I could hit keys at the piano, but I could never remember ("have a feel for") the music I was playing. And I lost interest.
Another example is chess. I always sucked at chess, because I can't remember opening books, prior games and other useful information. It both means I can't progress in games, and that I'm not fun in a social chess setting.
What I'm proposing, I guess, is that IQ tests are really judging what type of information you can remember/recall. E.g. extending number series: you can solve them either by remembering rules you can apply, or instances of them you've seen before. If you can remember instances, you'll be named a prodigy because you can do it much faster than learning rules. If you tend to remember instances, you'll probably want to continue finding examples to learn, as a dopamine hit.
EPWN3D
Respectfully no, I don't think they are. They are talents. Your neighbor probably cannot not teach someone how to look at an equation and visualize a plot of it. That's just a thing he can do.
Bill Clinton probably cannot teach someone how to remember a rando's name after several years. That's just something he can do that makes him a uniquely talented politician.
Intelligence is intended to be a metric, and that requires commonality. Mostly anyone can be taught to read, write, do basic math, recognize patterns in shapes, etc. That's why these things are used as indicators of intelligence. Saying that someone is more intelligent than everyone else because they possess a certain non-replicable talent undermines the entire endeavor.
prerok
I disagree. There is no qualifier in intelligence that would require the person to be able to transfer it to another person.
In fact, I think the IQ as a metric is fundamentally flawed. IIRC it rose from detecting learning disabilities, like if you are 10 years old and still don't know that if I write 1, 2, 3 and you don't guess the next number is 4, then there is likely something wrong and your IQ is below average. So, they took that test and wrote new, more advanced questions, like 2, 3, 5, what's next? Could be 7 (primes) or could be 8 (Fibonacci).
And the problem with that is that they really rely on what you are learning or not learning in school. The math example above might be passable but what about similes in literacy or what sides does this dodecahedron disassemble to. The former where somewhat taught in my school but not well and the latter... well, the children are supposed to experiment with shapes in their own time, apparently.
I do agree with you that it's some metric, but it just measures how well the material in the IQ test is taught in schools, it does not measure a person's intelligence, which is IMHO an innate quality.
threatofrain
The idea of general intelligence arose out of correlation between test performances, and that there was a more "general" capability that people were tapping into to solve the plethora of different intelligence tests.
prisenco
I've become convinced of the idea that everything on some level is a matter of practice, and that effortless practice comes from enjoyment.
Maybe your neighbor can visualize graphs because he practiced visualizing graphs a lot. And he did that because he found an odd joy in it that others didn't.
Bill Clinton knew he wanted to be in politics at a very early age by all accounts. And considering he didn't exactly come from a powerful political family with a clear path forward, it's likely he decided to practice remembering people and by the time anyone could comment on it, he'd had decades of practice. But honestly maybe he just had fun seeing what he could remember about the people he met and that turned it into a fun game for him that paid off big.
When I was young, I played a game where I would count the letters in the words people were saying to determine whether the total sentence was divisible by three. So for a time in my life I was unusually good at doing modulo division on people's conversation.
I've since lost that ability (not much use to it), but I wasn't some savant. My mother worked early shifts at the hospital and would drive me and my siblings to my grandmother's house, and the fog and dew would cling to the windshield regardless of whether it was raining so the windshield wipers would be on. And I would sync the voices on the AM radio to the three positions of the wipers, left, middle, right, left, middle, right. It was a game to pass the time because I couldn't fall back asleep after being woken up. By the time I could do it as a party trick, I'd had hundred of hours of practice.
brookst
I think you underestimate the role that hardware acceleration plays. Take the Clinton example — about 2% of the population has prosopagnosia (“face blindness”), where no amount of practice will improve recognition because of defects in the fusiform gyrus.
But! Just like some percentage don’t have the neurons to be great at recognizing people, there are “super recognizers” where the fusiform gyrus overperforms. fMRI studies have shown these people have more activity in this area.
Some skills, like your letter counting, probably do not have an evolved area of the brain and are purely cognitive, so perhaps more amenable to training. But even then there may be aspects that are hardware accelerated with some intrinsic genetic predisposition to be good / bad at.
somenameforme
> fMRI studies have shown these people have more activity in this area.
Neuroscience is paradoxically a poor indicator for how the brain works in response to stimuli. The reason is that what you do literally reshapes your brain and how you respond to things. For a common example a chess master and a chess amateur use different parts of then brain when analyzing positions. Here [1] is some random study on it - there's a bunch. It's well known and not controversial. But the thing is, that chess master was also once a chess amateur. And so practice literally changes how your brain responds to various stimuli to the point of complete shifting the area used for processing.
So showing that [people of this group] have [this distinct brain activity] in response to [whatever] doesn't really say anything at all, which it makes for endless easy mode studies for now. Hence people exposed to [video games, porn, gender dysmorphia, etc] all 'change the brain' and/or have 'different brains than normal.' Well duh, now so do you after reading this paragraph.
That said, I obviously would not dispute that people are inherently different, but I think the most obvious ways to try to demonstrate that are irreparably flawed.
[1] - https://www.scientificamerican.com/article/brain-study-shows...
prisenco
Sure, maybe. Some people might have some level of cognitive multiplier, but most of us aren't Ramanujan, most of us are some gradient of normal. Putting emphasis on practice and hard work and finding joy in getting good at something would pay greater dividends generally than trying to pinpoint a perfectly quantifiable intelligence.
leetbulb
Well stated. These abilities amplify the effectiveness of mental models we use to build useful systems i.e. they're force multipliers in discrete domains. People with these unique cognitive strengths become what we know as "principals." They can push beyond normal limitations in their areas.
Further, an organization that spots this dynamic and builds around it gains a real edge, assuming the company culture supports it.
nycdatasci
Excellent points. I’d add that the idea of intelligence diversity extends beyond people. We have extremely biased definitions intelligence that are very conveniently aligned with human strengths.
Ferret7446
The elephant in the room is that whatever IQ is measuring, it is strongly correlated with all of the things people would generally agree correlates with intelligence. It is a strong predictor for what is generally considered success: financial, health, achievements, career. It is a strong predictor for how fast people learn things or become competent in roles generally associated with intelligence.
You can nitpick it, you can be right, but you are also preaching to the rocks and trees. The fact is people with low IQs are going to struggle more and people with high IQs do not need to struggle as much, in all areas of life. The exact number, or accuracy, or whether it is really intelligence does not change the reality.
kstenerud
The author spoke to this:
"I would definitely never say something like “IQ doesn’t matter at all.” I wouldn’t even say “IQ is unimportant.” I think it is important, in that it’s one of the only measurements we have that does an okay job at capturing intelligence, in that it’s not too bad at this when it comes to the center of the distribution, although it gets increasingly bad at it at the tails.
And, from a practical perspective, there is a sense in which I’m actually very pro-IQ tests!"
randomNumber7
Yet I have high IQ and life is a struggle.
kbrkbr
"Throwing a die, a number higher than 1 is very likely." - "And yet I threw a 1."
That is no contradiction.
That being said I hope you can overcome the issues that make your life a struggle.
Arthanos
The ZIP code you're born in is a stronger predictor of all those things than IQ.
oidar
Do you have a citation for that?
threatofrain
And there are other people living in the same zip code as you. A single family may even have more than one child living under the same roof. Naturally the question will arise anyway as to why there are differences in learning speed or fluency.
jxjnskkzxxhx
You're just parroting a meme you saw somewhere, sounded reasonable, and never bothered to look for a reference.
superposeur
In any field, what it even means to be good morphs as you go up in skill level. Non mathematicians know only about arithmetic so they often imagine that mathematicians must be really really good at arithmetic. But this isn't so. Likewise, non musicians think what must make a great musician is perfect pitch. But some of the greatest musicians in history didn’t have it while many mediocre ones do. Similarly, non chess players think GMs must be good at calculating zillions of moves in advance, but apparently they only calculate a small set of moves, which somehow are the right ones.
To take an example cited in the article, Einstein was so far up there that it’s nearly impossible for a non physicist to even understand what he was so good at — crude measures like high school grades or “IQ” barely scratch the surface of the skill that he was a genius at.
Now, perfect pitch does modestly correlate with musical ability, mathematicians are better than average at arithmetic, GMs do calculate more moves than the average shmo, and Einstein got much better than average grades (after all he was accepted at ETH). But that’s all, modest correlations.
There is such a thing as talent in music, mathematics, etc. but it isn’t something a psychologist standing outside these domains would ever be able to devise a test for.
magicalhippo
When I was a teen I took a IQ test that was supposed to measure up to 160. It had quite a few questions but due to the scaling, just getting one question correct made it declare you had like 120 IQ or so.
Turned out I did indeed get one of the questions right, a fairly straight forward math question. Decided the test was probably bogus, just seemed dumb to conclude IQ based on a single correct answer.
Growing up some more I realized intelligence is complicated and multifaceted.
That I can solve some puzzles and usually think things through doesn't preclude me from being dumb as a brick in other situations.
wongarsu
When I did a psychologist-administered IQ test it took like 3 hours and resulted in four scores on different metrics. For example I had a pretty average 103 IQ on language, but about 120 IQ on logical thinking. In the end you get a total score from the average of the subcategories.
All IQ tests are pretty incomplete and a bit stupid, but some are less stupid than others
singularity2001
on an old computer I got an IQ score of 210 just by memorizing all the results still felt pretty good
milesrout
That isn't an IQ test. You can't do them on computers. They are assessed in person by professionals. You can't practice because the test aren't publicly available.
etempleton
What people typically refer to as an IQ test was not designed nor is it considered to be an accurate test of genius. It is not meant to mean anything if someone scores particularly well. The reason these tests are administered is to understand if there is perhaps a learning disability. If you score low overall, or low in a particular area that is then used as a diagnostic tool. For example, a normal score overall with a markedly low working memory score might point to a number of diagnoses.
Yes, scoring high is impressive, but is pretty meaningless in terms of outcomes beyond a point. Most IQ test scores are just made up. And there are plenty of self proclaimed genius IQ people who seem to be completely unremarkable beyond being a good test taker.
elcritch
> The reason these tests are administered is to understand if there is perhaps a learning disability. If you score low overall, or low in a particular area that is then used as a diagnostic tool. For example, a normal score overall with a markedly low working memory score might point to a number of diagnoses.
That's how I discovered my learning disability. Sort of ironic given that I'm a voracious learner and have become quite proficient in a couple of fields. However I'm a relatively slow learner, sometimes have quite dumb moments, and struggle with tests or being put on the spot. Usually when I do grasp a topic it's at a deep intuitive level.
Still a psychologist gave me a proper IQ exam after I emailed him saying I felt something was off in my brain. Testing took half a day. When I went back for results my score was close to the average for college graduates. Yet the phycologist asked me when I was going to get a PhD.
That confused me as my core didn't seem impressive. Then he said I probably had ADHD. Like many 80's kids I'd always sorta thought ADHD was mostly made up. Well he explained the results to me, in particular my low working memory compared to the rest of my scores on other subtests. My IQ dropped 40-50 points on some particular subscores which correlates very highly with ADHD.
By the end of that meeting I'd begun to realize, that it made a lot of sense with my struggles over the years. So yes, IQ tests are very helpful for diagnosis. I later got an official diagnosis at university.
I'm glad the psychologist offered me that IQ test. I hope others consider IQ testing if you're "smart but stuck".
kstenerud
IQ scores in the "sweet spot" (80-120) correlate quite well with academic achievement and income.
MisterKent
No mention of standard deviation? Just by the fact the IQ is your place on a normal curve, with a 15 point standard deviation, there are at least 100K+ people in the world with an IQ of 160.
The smartest 0.0032% of the population.
kstenerud
The point of the article was that above around 120 or so (and also at the low tail), the tests become very unreliable (due to a chicken-and-egg problem of finding people to calibrate the tests on). So an IQ score of 160 is highly suspect, and subject to even 2-digit variance.
nradov
Does intelligence follow the standard distribution?
JumpCrisscross
> Does intelligence follow the standard distribution?
Intelligence may or may not. IQ, by definition, does.
lifthrasiir
IQ as an intelligence rank does indeed follow the standard distribution, but in practice IQ is estimated by tests. I believe the GP wanted to ask about a distribution of such "estimated" IQs.
ks2048
So IQ can be negative? (Normal distribution has non zero probability mass for all real numbers)
nick__m
I am pretty sure that the central limit theorem applies to a sample size as big as all living humans.
pfedak
That isn't at all what the central limit theorem says. The whole point is it holds independent of the actual shape of distribution of the population. You could use the same argument to say social security numbers are normally distributed.
One way to explain things like height being normally distributed is that there are a bunch of independent factors which contribute, and the central limit theorem applied to those factors would then suggest the observed variable looking normal-ish.
ziofill
Average intelligence does :)
HDThoreaun
IQ does, by definition.
mod50ack
As an aside — intelligence is not the same thing as accomplishment. A lot of people do well in school (and do well on IQ tests) and do not achieve anything of note. The people we remember are those who achieve great things, not those who do well on puzzle-solving tests. Those aren't the same thing, and this is probably why Stephen Hawking said that caring about IQ is for losers.
Hawking was undoubtedly a smart guy, but that fact alone did not make his career. He did a lot of hard work in a field he was passionate about. Why would you tell Stephen Hawking how good you are at solving puzzles — why would he care about that? I'm sure he would have found news of some finding relevant to his research interests much more compelling.
Now, are IQ-type tests useful? Yes. They are quite good to administer in school to gauge people's reasoning abilities — to a certain point. The point of the tests was never to rank the smartest people, and to think about these edge cases — the ones tests are worst at measuring — is pointless. There are better things to concern yourself with; life is not an IQ leaderboard.
lifthrasiir
The first ever intelligence test (Binet-Simon) was designed partly in order to identify children who have intellectual disability and give a better treatment for them. The test itself was crude, but I believe that should be the only meaningful use of IQ tests because overly low or high IQ does predict certain kinds of complications. Otherwise they are easy to game, they only measure a particular slice of human intelligence, and their error bar is large enough that even a unit difference in SD can be not meaningful [1]. IQ is just a meaningless number that whoever have a higher one tends to have unjustified superiority over others.
[1] Depends on the particular test of course, but I can safely guess that +2 SD and +2.5 SD are not statistically distinguished by most tests.
mod50ack
Caring too much about IQ is not exempt from Goodhart's law.
modeless
When I first started using OTPs from Google Authenticator I found it difficult to remember random 6 digit numbers in one glance. Today after doing it almost daily for years and years I find it easy. Almost anything can be improved with training, including IQ test tasks, contrary to what test makers say. But I don't believe that such training improves general intelligence.
skirmish
I still find it hard to remember 8 digit OTPs, and work around that by repeatedly "singing" them in my mind while I type them. I can easily repeat those 8 meaningless words in my mind, and then decipher only the next digit I need to type.
throwawayk7h
I read that studying for an IQ test generally causes a gain of about 6 points at most. (Which is quite a bit.)
defrost
Not only is IQ a poor measure that's perhaps best at capturing near average performance, "intelligence" is an attribute that's not readily ranked in a linear manner as height can be.
JumpCrisscross
> "intelligence" is an attribute that's not readily ranked in a linear manner as height can be
IQ isn't a linear scale [1][2]. (It is ordinal.)
defrost
Who said that it was?
JumpCrisscross
> Who said that it was?
You said "'intelligence' is an attribute that's not readily ranked in a linear manner" while complaining about IQ. IQ doesn't rank intelligence linearly.
trts
generally intelligent people tend to perform well on intelligence tests
people who perform well on intelligence tests may not present as generally intelligent
Perenti
IQ measures how good you are at IQ tests. I'm brilliant at those tests, I consistently score 5+ SD above 100.
I'm also an idiot, and pretty much an asshole. IQ itself puts no beans on the table, and has little to no relationship to how good a human being you are.
adwn
> I'm brilliant at those tests, I consistently score 5+ SD above 100.
Are those professional tests supervised by a psychologist? Or rather "IQ tests" on websites that vastly inflate results to get you to share the results on social media? Both the "consistently" and the "5+ SD" (equivalent to an IQ of 175) make me think it's the latter.
Perenti
I'm talking about sit down in a room with a psychologist for hours, lots of questions, coloured blocks, tricky pictures, with stop watches etc. Reciting lists of numbers backwards and forwards, the _real_ tests, not pop quizes.
milesrout
How are you managing to convince psychologists to repeatedly test you?
Judgmentality
> I consistently score 5+ SD above 100
You consistently score 175+ on IQ tests?
Perenti
Yes. IQ tests are easy. Life, now that's a real challenge.
zem
can we put the "your" back in the title? it's currently unfairly ungrammatical.
raldi
I’m curious what could be the intent behind that rewrite rule.
zzo38computer
I will define "IQ" as what your score will be on a ideal IQ test (rather than an actual IQ test, although then you will have to define how the "ideal IQ test" is working, which is presumably like an actual IQ test if made and administered in the ideal way, however that might be). It is correlated with intelligence (which cannot be measured by a single number), but is not the same thing as intelligence. (I think it also not the same thing as genius, although that is also correlated with high intelligence, since intelligence is one of the things to be genius but is not the only part of it.)
I’ve often wondered how IQ correlates with the ability to visualize or recall audio of certain things.
For example, my neighbor in California - he had an ability to look at an equation and actually visualize the plot of it like a graphing calculator. That is certainly intelligence.
I got into asking people how they visualized for a while because I myself don’t. But then I’m apparently good at knowing if I’ve seen somebody or not, no matter how short of time it was I’ve seen them. I’m sorta the same way with having been a place, usually knowing how to get back there, without actually seeing it. Sometimes I wonder if I do that “similar” thing with everything, including ideas.
My stepdad once said he met Clinton briefly once and then Clinton saw him many years later at a Razorback game and he instantly remembered my dad’s name. That’s a level of intelligence.
And if somebody has perfect pitch, or can listen to a song and play it back in the piano, then that’s also intelligence of a certain type, maybe.
So what makes an IQ test be able to measure such variety when some have totally different modes of thinking?