Admitting a bias is the first step to overcoming it, so I’ll admit it: I have a huge bias against growth mindset.

(if you’re not familiar with it, growth mindset is the belief that people who believe ability doesn’t matter and only effort determines success are more resilient, skillful, hard-working, perseverant in the face of failure, and better-in-a-bunch-of-other-ways than people who emphasize the importance of ability. Therefore, we can make everyone better off by telling them ability doesn’t matter and only hard work does. More on Wikipedia here).

See, I can sometimes be contrarian, and growth mindset is pretty much the only idea from social psychology that is universally beloved. If I try to search for criticism of growth mindset, I am buried in the Google-shadow of people raving about how wonderful a discovery it is and how we all need more of it. Google “growth mindset debunked” and you just get a bunch of articles talking about how growth mindset debunked all the other inferior ideas before it was discovered. Google “growth mindset publication bias”, and you just get a bunch of articles on how we need to keep a growth mindset about fighting publication bias.

It’s unnatural, is what it is. A popular psychological finding that doesn’t have gruff people dismissing it as a fad? That doesn’t have politicians condemning it as a feel-good justification for everything wrong with society? That doesn’t have a host of smarmy researchers saying that what, you still believe that, didn’t you know it failed to replicate and has since been entirely superseded by a new study out of Belarus? I’m not saying Carol Dweck has definitely made a pact with the Devil, I’m just saying I don’t have a good alternative explanation.

Which brings me to the second reason I’m biased against it. Good research shows that inborn ability (including but not limited to IQ) matters a lot, and that the popular prejudice that people who fail just weren’t trying hard enough is both wrong and harmful. Social psychology has been, um, very enthusiastic about denying that result. If all growth mindset did was continue to deny it, then it would be unexceptional.

But growth mindset goes further. It’s not (just?) that ability doesn’t matter. It’s that belief that ability might matter is precisely what makes people fail. People who believe ability matters will refuse to work hard, will avoid challenges, will become “helpless” in the face of pressure, will hate learning as a matter of principle, will refuse to work hard, will become blustery and defensive about their “brilliance”, will lie to people and hide their failures, and will drop out of school and turn to drugs (really)! People who believe that anyone can succeed if they try hard enough will be successful, well-adjusted, and treat life as a series of challenging adventures. It all strikes a curmudgeon like me as just about the thickest morality tale since Pilgrim’s Progress, and as just about the most convenient explanatory coup since “the reason psychic powers don’t work on you is because you’re a skeptic!”

Which brings me to the third reason I’m biased against it. It is right smack in the middle of a bunch of fields that have all started seeming a little dubious recently. Most of the growth mindset experiments have used priming to get people in an effort-focused or an ability-focused state of mind, but recent priming experiments have famously failed to replicate and cast doubt on the entire field. And growth mindset has an obvious relationship to stereotype threat, which has also started seeming very shaky recently.

So I have every reason to be both suspicious of and negatively disposed toward growth mindset. Which makes it appalling that the studies are so damn good.

Consider Dweck and Mueller 1998, one of the key studies in the area. 128 fifth-graders were asked to do various puzzles. First they did some easy ones and universally succeeded. The researchers praised them as follows:

All children were told that they had performed well on this problem set: “Wow, you did very well on these problems. You got [number of problems] right. That’s a really high score!” No matter what their actual score, all children were told that they had solved at least 80% of the problems that they answered.

Some children were praised for their ability after the initial positive feedback: “You must be smart at these problems.” Some children were praised for their effort after the initial positive feedback: “You must have worked hard at these problems.” The remaining children were in the control condition and received no additional feedback.

This is a nothing intervention, the tiniest ghost of an intervention. The experiment had previously involved all sorts of complicated directions and tasks, I get the impression they were in the lab for at least a half hour, and the experimental intervention is changing three short words in the middle of a sentence.

And what happened? The children in the intelligence praise condition were much more likely to say at the end of the experiment that they thought intelligence was more important than effort (p < 0.001) than the children in the effort condition. When given the choice, 67% of the effort-condition children chose to set challenging learning-oriented goals, compared to only 8% (!) of the intelligence-condition. After a further trial in which the children were rigged to fail, children in the effort condition were much more likely to attribute their failure to not trying hard enough, and those in the intelligence condition to not being smart enough (p < 0.001). Children in the intelligence condition were much less likely to persevere on a difficult task than children in the effort condition (3.2 vs. 4.5 minutes, p < 0.001), enjoyed the activity less (p < 0.001) and did worse on future non-impossible problem sets (p...you get the picture). This was repeated in a bunch of subsequent studies by the same team among white students, black students, Hispanic students...you probably still get the picture. Or take An Analysis Of Learned Helplessness. Dweck has used a test called the IAR to separate children out into those who think effort is more important (“mastery-oriented”) and those who think ability is more important (“helpless”). Then she gave all of them impossible problems and watched them squirm – or, more fomally, tested how long the two groups continued working on them effectively. She found extremely strong results – of the 30 subjects in each group, 11 of the mastery-oriented tried harder after failure, compared to 0 helpless. 21 of the helpless children stopped trying hard after failure, compared to only 4 mastery-oriented. She described the mastery-oriented children as saying things like “I love a challenge,” and the helpless children begging to be allowed to stop.

This study is really weird. Everything is like 100% in one group versus 0% in another group. Either something is really wrong here, or this one little test that separates mastery-oriented from helpless children constantly produces the strongest effects in all of psychology and is never wrong. None of the children whose test responses indicated that they thought ability was important to success ever monitored their own progress – not one – while over 95% of the children who said they thought effort was more important did. None of them ever expressed a positive statement about their own progress, while over two-thirds of the children who thought effort was more important did.

Normally I would assume these results are falsified, but I have looked for all of the usual ways of falsifying results and I can’t find any. Also, the boldest falsifier in the world wouldn’t have the courage to put down numbers like these. And a meta-analysis of all growth mindset studies finds more modest, but still consistent, effects, and only a little bit of publication bias.

So – is growth mindset the one concept in psychology which throws up gigantic effect sizes and always works? Or did Carol Dweck really, honest-to-goodness, make a pact with the Devil in which she offered her eternal soul in exchange for spectacular study results?

I don’t know. But here are a few things that predispose me towards the latter explanation. A warning – I am way out of my league here and post this only hoping it will spark further discussion.

II.

The first thing that bothers me is the history.

I’ve been trying really hard to trace its origin story, but it is pretty convoluted. It seems to have grown out of a couple of studies Carol Dweck and a few collaborators did in the seventies. But these studies generally found that a belief in innate ability was a positive factor alongside belief in growth mindset, with the problem children being the ones who attributed their success or failure to bad luck, or to external factors like the tests being rigged (which, by the way, they always were).

A good example of this genre is Learned Helplessness And Reinforcement Responsibility In Children. Its abstract describes the finding as: “Subjects who showed the largest performance decrements were those who took less personal responsibility for the outcomes of their actions…and who, when they did accept responsibility, attributed success and failure to presence or absence of ability rather than to expenditure of effort.”

But that seems like a somewhat loaded way of interpreting this table:

As you can see, the “persistent” children (the ones who kept going in the face of failure) had stronger belief in the role of ability in their successes (I+a) and failures (I-a) than the “helpless” children (the ones who gave up in the face of failure)! These don’t achieve statistical significance in this n = 10 study, but they do repeat across all four combinations of success x gender tested. The real finding of the study was that children who attributed their success or failure to any stable factor, be it effort or ability, did better than those who did not.

Likewise, in The Role Of Expectations And Attributions, Dweck describes her findings as “persistent and helpless children do not differ in the degree to which they attribute success to ability”. When you actually look at the paper, this is another case of the persistent children actually having a higher belief in the importance of ability, which fails to achieve statistical significance because the study is on a grand total of twelve children.

(I should say something else about this study. Dweck compared two interventions to make children less helpless and better at dealing with failure. In the first, she gave them a lot of easy problems which they inevitably succeeded on and felt smart about. In the second, she gave them difficult problems they were bound to fail, then told them it was because they weren’t working hard enough. Finally, both groups were challenged with the difficult bound-to-fail problems to see how hard they tried on them. The children who had been given the impossible problems before did better than the ones who felt smart because they’d only gotten easy problems. Dweck interpreted this to prove that telling children to work hard made them less helpless. To me the obvious conclusion is that children who are used to failing get less flustered when presented with impossible material than children who have artificially been made to succeed every moment until now.)

Then there’s there’s this, a preliminary to the second study I mentioned in Part I. Does it show the mastery-oriented children outperforming the helpless children on every measure. Yeah. But listen to this part from the discussion section:

The results revealed striking differences both in the pattern of performance and in the nature of the verbalizations made by helpless and mastery-oriented children following failure. It was particularly noteworthy that while the helpless children made the expected attributions to uncontrollable factors, the mastery-oriented children did not offer explanations for their failures

.
But if you look at the data, this doesn’t seem right.

Mastery-oriented children were about six times more likely to attribute their failures to the most uncontrollable factor of all – bad luck. They were also about six times more likely to attribute their failures to the task “not being fair”. This contradicts every previous study, including Dweck’s own. The whole field of attribution theory, which is intensely studied and which Dweck cites approvingly, says that attributing things to luck is a bad idea and attributing them to ability is, even if not as good as effort, pretty good. But Dweck finds that the kids who used ability attributions universally crashed and bomb, and the kids who attribute things to luck or the world being unfair do great.

It might not be fair for me to pick on these couple of small studies in particular when there’s so much out there, but the fact is that these are the first, and a lot of the reviews cite only these and a few theses which as far as I know were never published. So this is what I’ve got. And from what I’ve got, I find that until about 1980, every study including Dweck’s found that belief in ability was a protective factor. Suddenly this disappeared and was replaced with it being a toxic plague. What happened? I don’t know.

III.

The second thing that bothers me is the longitudinal view.

So you have your helpless, fixed-mindset, believe-in-innate-ability children. According to Dweck, they “…are so concerned with being and looking talented that they never realize their full potential. In a fixed mindset, the cardinal rule is to look talented at all costs. The second rule is don’t work too hard or practive too much…having to work casts doubt on your ability. The third rule is, when faced with setbacks, run away. They say things like ‘I would try to cheat on the next test’. They make excuses, they blame others, they make themselves feel better by looking down on those who have done worse.”

These people sound like total losers, and it’s clear Dweck endorses this reading:

“Almost every great athlete – Michael Jordan, Jackie Joyner-Kersee, Tiger Woods…has had a growth mindset. Not one of these athletes rested on their talent…research has repeatedly shown that a growth mindset fosters a healthier attitude toward practice and learning, a hunger for feedback, a greater ability to deal setbacks, and significantly better performance over time…over time those with a growth mindset appear to gain the advantage and begin to outperform their peers with a fixed mindset.”

Man, it sure would be awkward if fixed mindset students generally did better than growth-mindset ones, wouldn’t it?

Aronson, Fried, and Good (2001) looks at first like just another stunning growth mindset study. They do a half-hour intervention to teach college students growth mindset and find they are still getting higher grades a couple of months later (an effect so shocking I wrote about it here). But one thing they do kind of as an afterthought is measure the students’ general level of growth mindset, as well as some measures of academic performance before the intervention.

People with high growth mindset had lower GPA (decent effect size but not statistically significant) and lower SAT scores (which was statistically significant).

The authors are obviously uncomfortable with this, but they propose that people who get low SAT scores just tell themselves ability doesn’t matter/exist in order to protect their self-esteem since they don’t seem to have much of it.

And okay, that’s probably true (a commenter makes the equally good point that smart people may coast on their native intelligence without ever applying effort, and so accurately describe their experience as ability mattering but effort not doing so).

But if Dweck is to be believed, people with growth mindset are amazing ubermenschen and people with fixed mindset are disgusting failures at everything who hate learning and give up immediately and try to cheat. In the real world, however big the effect is, it is totally swamped by this proposed “people with low SAT scores protect their self-esteem or whatever” effect.

The same study also notes the awkward result that blacks are more likely to believe intelligence is flexible and growth-mindset-y than whites, even though blacks do worse in school and even though half the reason people are pushing growth mindset is to try to explain minority underperformance.

This is not an isolated finding. For example, Furnham (2003) finds in a sample of students at University College London that mindset is not related to academic performance. I’ve been told there’s a study from Pennsylvania that shows the same thing, though I can’t find it.

If you look hard enough, you can even notice this in Dweck’s studies themselves. One little-remarked-upon feature of Dweck’s work is that the helpless children and the mastery-oriented children always start out performing at the same level. It’s only after Dweck stresses them out with a failure that the mastery-oriented children recover gracefully and the helpless children go into free-fall.

But these are fifth-graders! For the two groups of children to do equally well on the first set of problems means that from first through fourth grade, their “helpless” “fixed-mindset” work-hating nature hasn’t impaired their ability to learn the material to a fifth-grade level one bit! (In this study, the fixed mindset children actually start out doing better; I can’t find any studies where the growth mindset children do).

When it’s convenient for her argument, Dweck herself admits that:

Some of the brightest, most skilled individuals exhibit the maladaptive pattern. Thus, it cannot be said that it is simply those with weak skills or histories of failure who (appropriately) avoid difficult tasks or whose skills prove fragile in the face of difficulty.

But I don’t think she follows the full implication of this statement, that despite being doomed to failure by their fixed mindset, these people have become “the brightest and most skilled individuals”.

Indeed, there has recently been some growth mindset studies done on gifted students, at elite colleges, or in high-level athletics. All of these dutifully show that people with fixed mindset respond much worse to whatever random contrived situation the experimenters produce. But thus far nobody has pointed out that there seem to be about as many of these people at, say, Stanford as there are anywhere else. If growth mindset was so great, you would expect fixed mindset people at Stanford to be as rare as, say, people with < 100 IQ are at Stanford. Given that you will search in vain for the latter but have no trouble finding a bunch of the former for your study on how great growth mindset is, it sure looks like IQ is useful but growth mindset isn't. When people are in a psychology study, the fixed mindset individuals universally crash and bomb and display themselves to be totally incapable of learning or working hard. At every other moment, they seem to be doing equally well or better than their growth mindset peers. What's going on? I have no idea. IV.

The third thing that bothers me is Performance Deficits Following Failure, a study which manages to be quite interesting despite coming from a university in a city that very possibly doesn’t exist.

They use a procedure much like Dweck’s. They make children do some problems. Then they give them some impossible problems. Then they give them more problems, to see if they’ve developed “learned helplessness” from their failure on the impossible ones. Dweck’s theory predicts that the fixed-mindset children would and the growth mindset children wouldn’t. The Bielefeld team wasn’t testing growth mindset, but they indeed found that a bunch of children got flustered and stopped trying and did poorly from then on.

Then they repeated the experiment, but this time they made it look like no one would know how the children did. They told the kids they would be on teams, and the scores of everyone on their team would be combined before anybody saw it. The kids could fail as much as they wanted, and it would never reflect on them.

After that, children did exactly as well after failure as they had before. There was no sign of any decrease, or any “fixed mindset” group that suddenly gave up in order to protect their ego.

This doesn’t strike me as fully consistent with mindset theory. In mindset theory, people are acting based on their own deep-seated beliefs. Once a fixed mindset child fails, that’s it, she knows she’s Not Intelligent, there’s no helping it, all she can do is sabotage herself on the problems in order to protest a spiteful world that has failed to recognize her genius blah blah blah. Instead, there seems to be a very social role to these failures. The Bielefeld team describes it as “self-esteem protection”, but that doesn’t make much sense to me, since if they were worried about their self-esteem they could still be worried about it when no one else knew their performance.

To me it seems like some kind of interaction between self-esteem and other-esteem. Fixed mindset people get flustered when they have to fail publicly in front of scientists. This doesn’t seem like an unreasonable problem to have. A more interesting question is why it’s correlated with belief in innate ability.

Suppose that the difference in “people who talk up innate ability” and “people who talk up hard work” maps onto a bigger distinction. Some people really want to succeed at a task; other people just care about about clocking in, going through the motions, and saying “I did what I could”.

Put the first group in front of an authoritative-looking scientist, tell them to solve a problem, and make sure they can’t. They’re going to view this as a major humiliation – they were supposed to get a result, and couldn’t. They’ll get very anxious, and of course anxiety impedes performance.

Put the second group in front of an authoritative-looking scientist, and they’ll notice that if they write some stuff that looks vaguely relevant for a few minutes until the scientist calls time, then whatever, they can say they tried and no one can bother them about it. They do exactly this, then demand an ‘A’ for effort. At no point do they experience any anxiety, so their performance isn’t impeded.

Put both groups on their own in private, and neither feels any humiliation, and they both do about equally well.

Now put them in real life. The success-oriented group will investigate how to study most effectively; the busywork-oriented group will try to figure out how many hours of studying they have to put in before other people won’t blame them if they fail, then put in exactly that amount. You’ll find the success-oriented group doing a bit better in school, even though they fail miserably in Dweck-style experiments.

And if an experimenter praises children for working hard, it will make them believe that all the experimenter cares about is their effort. Next problem, when the experimenter poses an impossible question, the child will beat their head against it for no reason, since that’s apparently what the experimenter wants. But if the experimenter praises a child’s ability, then the child will feel like the experimenter really wants them to correctly solve the questions. When the next question proves unsolveable, the child will admit it and expect the experimenter to be disappointed.

I doubt that this is the real phenomenon behind growth mindset, simply because it flatters my own prejudices in much the same way mindset theory flatters everyone else’s. But I think it shows there are a lot of different narratives we could put in this space, all of which would be able to explain some of the experimental results.

V.

I want to end by correcting a very important mistake about growth mindset that Dweck mostly avoids but which her partisans constantly commit egregiously. Take this article, Why A Growth Mindset Is The Only Way To Learn:

[Some people think] you’ll always have a set IQ. You’re only qualified for the career you majored in. You’ll never be any better at playing soccer or dating or taking risks. Your life and character are as certain as a map. The problem is, this mindset will make you complacent, rob your self-esteem and bring meaningful education to a halt.

In short, it’s an intellectual disease and patently untrue.

The article goes on to show how growth mindset proves talent is “a myth”, a claim repeated by growth mindset cheerleader articles like Debunking The Genius Myth and The Learning Myth: Why I’ll Never Tell My Son He’s Smart and this woman who says we need to debunk the idea of innate talent.

Suppose everything I said in parts I – IV was wrong, and growth mindset is 100% true exactly as written.

This still would not provide an iota of evidence against the idea that innate talent / IQ / whatever is by far the most important factor determining success.

Consider. We know from countless studies that strong religious belief increases your life expectancy, makes you happier, reduces your risk of depression and reduces crime. Clearly believing in, say, Christianity has lots of useful benefits. But no one would dare argue that proves Christianity true. It doesn’t even imply it.

Likewise, mindset theory suggests that believing intelligence to be mostly malleable has lots of useful benefits. That doesn’t mean intelligence really is mostly malleable. Consider, if you will, my horrible graph:

Suppose this is one of Dweck’s experiments on three children. Each has a different level of innate talent, represented by point 1. After they get a growth mindset and have the right attitude and practice a lot, they make it to point 2.

Two things are simultaneously true of this model. First, all of Dweck’s experiments will come out exactly as they did in the real world. Children who adopt a growth mindset and try hard and practice will do better than children who don’t. If many of them are aggregated into groups, the growth mindset group will on average do better than the ability-focused group. Intelligence is flexible, and if you don’t bother practicing than you fail to realize your full potential.

Second, the vast majority of difference between individuals is due to different levels of innate talent. Alice, no matter how hard she practices, will never be as good as Bob. Bob, if he practices very hard, will become better than Carol was at the start, but never as good as Carol if she practices as hard as Bob does. The difference between Alice and Carol is a vast, unbridgeable gap which growth mindset has nothing whatsoever to say about.

Here is a graph which is less terrible because it was not made by me. I have taken it from one of the two other sources I have found on the entire Internet that don’t like growth mindset:

We can argue all day about whether poor students do worse because they have bad health, because they have bad genes, because they have bad upbringings, or because society is fixed against them. We have argued about that all day before here, and it’s been pretty interesting.

But in this case it doesn’t matter. If the only thing that affects success is how much effort you put in, poor kids seem to be putting in a heck of a lot less effort in a surprisingly linear way. But the smart money’s not on that theory.

A rare point of agreement between hard biodeterminists and hard socialists is that telling kids that they’re failing because they just don’t have the right work ethic is a crappy thing to do. It’s usually false and it will make them feel terrible. Behavioral genetics studies show pretty clearly that at least 50% of success at academics and sports is genetic; various sociologists have put a lot of work into proving that your position in a biased society covers a pretty big portion of the remainder. If somebody who was born with the dice stacked against them works very hard, then they might find themselves at A2 above. To deny this in favor of a “everything is about how hard you work” is to offend the sensibilities of sensible people on the left and right alike.

Go back to that 1975 paper above on “Role Of Expectations And Attributions” and look more closely at the proposed intervention to help these poor fixed mindset students:

Twelve extremely helpless children were identified [and tested on how many math problems they could solve in a certain amount of time]…the criterion number was set one above the number he was generally able to complete within the time limit. On these trials, he was stopped one or two problems short of criterion, his performance was compared to the criterion number required, and experimenter verbally attributed the failure to insufficient effort.

So basically, you take the most vulnerable people, set them tasks you know they’ll fail at, then lecture them about how they only failed because of insufficient effort.

Imagine a boot stamping on a human face forever, saying “YOUR PROBLEM IS THAT YOU’RE JUST NOT TRYING NOT TO BE STAMPED ON HARD ENOUGH”.

And maybe this is worth it, if it builds a growth mindset that allows the child to be more successful in school, sports, and in the rest of her life. But you’re not “debunking the myth of genius”. Genius remains super-important, just like conscientiousness and wealth and health and privilege and everything else. No, you’re telling a Noble Lie to the children because you think it’s useful. You can make it palatable by saying “Well, we’re not denying reality, we’re just selectively emphasizing certain parts of reality, but in the end that’s what you’re doing. If you can square that with your moral system, go ahead.

But I remain agnostic. There are some really good – diabolically good? – studies showing that it works in certain lab situations. There’s a lot of excellent research behind it and a lot of brilliant people giving it their support. But there are also other studies showing that it has no long-term real-world effects that we can measure, and others that might (or might not?) contradict its predictions in other ways. I have only the barest of ideas how to square those facts, and I look forward to hearing from anyone who has more.

I.

Last week I challenged a bad study about innate ability, and in the process I accidentally made a few people feel depressed and worthless. Yesterday I tried to resolve that issue, and in the process I accidentally made a few people feel like effort doesn’t matter and there’s no point in trying hard at things. For example:

If I had read this post back in those days, before I had signed up for my class— well, I suppose I never would have signed up. I would have accepted that I did not understand mathematics and that I never could understand mathematics and leave it at that.

So I guess we have to continue our game of Crippling Self-Doubt Whack-A-Mole. Goodness only knows what horrible lesson people will draw from this post. Please, if you interpret it to mean you should light yourself on fire or something, check with me first, okay?

II.

The average eminent theoretical physicist has an IQ of 150-160. The average NBA player has a height of 6’7. Both of these are a little over three standard deviations above their respective mean. Since z-scores are magic and let us compare unlike domains, we conclude that eminent theoretical physicists are about as smart as pro basketball players are tall.

Any time people talk about intelligence, height is a natural sanity check. It’s another strongly heritable polygenic trait which is nevertheless susceptible to environmental influences, and which varies in a normal distribution across the population – but which has yet to accrete the same kind of cloud of confusion around it that IQ has.

So let’s see what we can learn from the height distribution of every player in the NBA: (source)

The first thing we notice is that nobody shorter than average (ie 5’9) gets in. This is not true eternally and forever – Wikipedia’s List Of Shortest NBA Players names 25 NBA players in history (out of maybe 5000 total) who were below average stature. The all-time record holder is Tyrone “Muggsy” Bogues at 5’3, who was actually pretty good (though not above exploiting his distinction for publicity: his autobiography is called In the Land of Giants). But Bogues was a spectacular outlier. If we’re going to stick to our resolution to use the histogram we’ve got, the contribution of the left half of the height bell curve is precisely zero.

We can do some very rough Fermi calculations for the next couple standard deviations.

34% of the US male population is 5’9 to 6’0, so about 54 million men. There are 5 NBA players in this band. So about one in every 11 million people of this height is in the NBA.

13.5% of the US male population is 6’0 to 6’3, so about 21 million men. There are 40 NBA players in this band. So about one in every 500,000 people of this height is in the NBA.

2.35% of the US male population is 6’3 to 6’6, so about 4 million men. There are 95 NBA players in this band. So about one in every 40,000 people of this height is in the NBA.

0.15% of the US male population is 6’6 to 6’9, so about 200,000 men. There are 130 NBA players in this band. So about one in every 1,500 people of this height is in the NBA.

0.003% of the US male population is 6’9 to 7’0, so about 5,000 men. There are 160 NBA players in this band. So about one in every 30 people of this height is in the NBA.

0.00002% of the US male population is 7’0 to 7’3, which corresponds to about 45 men. There are 40 NBA players in this band. So about 8 out of 9 people of this height are in the…wait, no, that can’t be right.

Sports Illustrated‘s Dan Diamond does a similar analysis and gets broadly similar results. But he adds several complicating factors. First, at this height people with very rare medical conditions and pituitary tumors start taking over from normal variation, so we lose our ability to derive a census a priori with pure math. Second of all, at this height talent scouts comb the world for suitably tall foreigners and import them, so we can no longer assume we’re drawing from the pool of tall Americans. And third of all, the same way you round up so that you’re 6’0 on OKCupid, NBA players round up so that they’re 7’0 on their official stats.

He concludes that most likely about 17% of seven-foot-tall young men in America are in the NBA. This might still be an overestimate, but is probably in the right ballpark. Forbes Magazine writes that Being Seven Feet Tall Is The Fastest Way To Get Rich In America and quotes a talent scout who says “that he’ll ‘check up on anyone over 7 feet that’s breathing.'” Given that a lot of people this height are unhealthy or uninterested, it might not be much of an exaggeration to say that if you’re 7’3 and have any interest in basketball, you’ve got better than even odds of going pro.

But why stop there? 0.000000001% of men are 7’3 to 7’6. Given the size of the American male population, there shouldn’t be a single person in the US with this height, though in practice a few people with endocrine abnormalities make the cut. There will, however, be a couple of healthy people of this height in the world population. There is one person currently in the NBA above 7’3, and he is a Tanzanian native discovered by talent scouts.

WE’VE GOT TO GO DEEPER TALLER! 0.00000000000001% of the population is 7’6 to 7’9. Statistically, there should not be a single man this tall in the entire world. No NBA players are currently this tall. But Yao Ming, who retired four years ago, was 7’6 exactly. He was the product of a Maoist breeding program specifically aimed at producing tall people to play basketball. You can break a lot of statistical laws if you have breeding programs and flexible ethics. Also if you have pituitary tumors, as the remainder of the List Of Tallest People reminds us. It looks like a little over half of the living people in this height band have played professional basketball at some point.

There are people taller than 7’9, but a lot of them have trouble standing without support, which somewhat decreases basketball ability.

On the plus side, you can do all sorts of awesome things, like take a picture standing next to the world’s shortest man, or…okay, pretty much just that.

III.

This offers us an opportunity to compare our intuitions about intellect to our intuitions about basketball. I recommend taking that opportunity. Basketball isn’t mysterious, it isn’t politicized, and it isn’t tied up with our notions of self-worth. As a result, our intuitions about basketball are crystal clear.

If somebody said “Height plays no role in your success as a basketball player,” we’d look at them funny. Literally nobody shorter than average is currently in the NBA. Your odds at average height are only one in ten million. But if you’re 7’0, you can pretty much walk right in.

But if somebody said “Well, success in basketball is clearly dependent on height, that means there’s no point in practicing, after all you’re not going to grow any taller,” we’d look at them funny too. I don’t know much about sports, but I predict a short guy who’s been playing for years can defeat an untrained giant every time. And of course if two giants go up against each other, experience will be decisive.

If someone found a 5’8 kid who really liked playing hoops in his backyard and was pretty good at it, and they told him “You’re a moron, you can never succeed at basketball because of your height, give up,” that person would be a jerk. The kid is enjoying himself, there’s no point in insulting him about it, and if his goal is just to end up as a college athlete or a minor leaguer, then with hard work he could certainly make it. Even if he didn’t, he might be able to apply what he’d learned to a related area like coaching or reporting.

But if that same kid wanted to go deeply into debt to attend a basketball training camp, and he’ll count it a failure if he achieves anything less than NBA superstardom, you should probably warn him that his hopes aren’t very realistic and that maybe he should lower his standards or pick a sport more suited to his body type or try another line of work entirely.

If somebody who was 6’6 complained that they’d never be able to beat the 7’0 players on the other side, we would tell them to brighten up. After all, Michael Jordan was “only” 6’6, and he’s maybe the greatest basketball player of all time, even though he often had to face off against people taller than himself.

But if a team made entirely of 6’6 players faced off against a team entirely of 7’0 players, and both of them were really motivated and practiced really hard and had great coaches, I know who I would bet on.

Most important, people all over the world innocently enjoy playing basketball, and they are right to do so. Everyone knows that taller players have an advantage, no one’s denying it. But at the levels most people play at, moderate height differences are surmountable by differences in training and technique, and large height differences are so rare as to not come up very often. Yes, the pro leagues are a different story. Yet somehow the 5’8 kid who scores a three-pointer still manages to feel good about himself. And we can be very impressed by Kobe Bryant and LeBron James, praise them for their determination and technique and competitive spirit, all while acknowledging that it’s not a coincidence that the one is 6’6 and the other 6’9 and neither of these is a normal-person height.

All of this is crystal clear when we’re talking about basketball. But as soon as we switch back to talking about intelligence, we’re shouting at each other: “YOU SAY RAMANUJAN WAS REALLY HARD-WORKING, BUT I THINK HE HAD HIGH IQ, SO THERE!” Or “WELL IF SUCCESS COMES FROM INNATE TALENT, I GUESS I’LL JUST NEVER STUDY AGAIN.”

Luke Muehlhauser liked to call his philosophy of religion “common sense atheism”, meaning that he wanted to treat the question of God with the same “common” reasoning that he used for every other question. If we don’t see a tiger in front yard, we don’t say “Since it’s impossible to prove a negative, I can at best be agnostic about the existence of a tiger”, we say “I guess there’s probably not a tiger.”

Likewise, if we can just apply the same common reasoning we use for normal everyday activities like basketball to the question of intelligence, we might find it’s not so complicated and scary after all.

Which is not to say that it’s not important. After all, the other analogy between intelligence and basketball talent is that they’re both skills we need to cultivate at the highest levels if we want to save the world when it is threatened by dangerous future technology we can barely comprehend

[still more on this subject later, but not immediately]

But while we’re figuring that out, can someone explain to me how polygenic inheritance works?

We have really strong evidence that intelligence is highly heritable – maybe 50% to 80%. But genome-wide association studies show very low contributions from any particular SNP:

The study found that the genetic markers with the strongest effects on educational attainment could each only explain two one-hundredths of a percentage point (0.02 percent). To put that figure into perspective, it is known from earlier research that the SNP with the largest effect on human height accounts for about 0.40 percent of the variation.

Combining the two million examined SNPs, the SSGAC researchers were able to explain about 2 percent of the variation in educational attainment across individuals, and anticipate that this figure will rise as larger samples become available.

Here’s my question. Things determined by a larger number of independent randomly varying processes should tend to vary less. Suppose at Casino A, you flip a coin and if it comes up heads they give you $5000 and if it comes up tails they give you nothing. But at Casino B, you flip 5000 coins, and get $1 for each coin that comes up heads and nothing for each coin that comes up tails.

Both games have the same minimum and maximum winnings ($0 and $5000), and both have the same average ($2500). But the variance will be very different. The winnings at Casino A will vary a lot; half the people will walk out with $5000 and the other half will walk out broke. The winnings at Casino B will vary surprisingly little: if I understand binomial distributions correctly, well under 1% of gamblers will walk out with less than $2400 or greater than $2600. Pretty much everyone goes home with something like $2500.

The more independent loci determine human intelligence, the less variation in human intelligence we expect to see relative to the total amount it is possible to vary. If the most important SNP explains 0.02%, then there are at least 5000 genes involved, which means, as in the example above, that less than 1% of people should differ more than 2% of total possible variability from the average.

But actually, people differ in intelligence a lot, and a lot of that difference seems to have a genetic component. This implies either that the total possible genetic variability in intelligence is huge – that the right genes could give you an IQ of 2500 or so, or that variability in intelligence isn’t simple and additive and random the way I’m modeling it here – or that I’m doing the math wrong, always a distinct possibility.

The easiest way to get out of this, other than accepting I am terrible at math and should be kept away from it, is to assume that lots of different genes for good intelligence are correlated. Maybe one population had reason to evolve high intelligence with lots of smart SNPs, and a second population didn’t. Then it would make complete sense that all the genes involved would co-vary. Unfortunately, there seems to be significant IQ variation within the same family, let alone within the same population group, so that fails pretty hard as an explanation.

Another possibility is to accept the whole mutational load idea – which allows for high correlations in goodness or badness of the entire genome. Unfortunately, this is the other idea in genetics which has been confusing me terribly over the past few months.

This has no trouble explaining correlations, but it does have trouble explaining why things aren’t more correlated. We should find the same people being very smart and very tall and very athletic and very healthy. I don’t doubt there are some correlations between these traits, but they don’t seem nearly as high as one might expect.

And the paternal age effect keeps being brought in to explain this, but I don’t get that one either. Suppose I have a kid at 60. My sperm and my DNA have had 60 years to accumulate deleterious mutations, so there’s more chance my kid will have low IQ or psychiatric disease or whatever. Fine. But suppose while I’m having a kid at 60, my twin brother has had a kid at age 20, and his kid had a kid at age 20, and his kid had a kid at age 20, so that his first great-grandchild is being born exactly the same time my first child is. Both my kid and his great-grandkid have had 60 years worth of cell dividings to accumulate mutations. Why should their risks of autism be any different just because his kid had those 60 years divided among three different people?

(does less frequent division of spermatogonia before puberty cause them to accumulate fewer mutations during that time? If so, shouldn’t three generations of people who have kids at 33 still accumulate 80% as many mutations as two generations of people who have kids at 50?)

I’ve heard evolution’s ability to eliminate people with bad genes used as an explanation here, but I don’t quite get it. For one thing, it seems unlikely that evolution can produce beneficial mutations at the same rate people accumulate deleterious mutations (10 per generation or so). And if we imagine two lineages of gradually deteriorating intelligence going on for 1000 years, the Lions That Selectively Eat Low-IQ People will have just as much opportunity to cull the members of the one that reproduces quickly as the one that reproduces slowly. And, if that were the explanation we should fail to see a paternal age effect in the absence of such lions (or be dysgenic as hell), but the effect has been demonstrated in our own society, which is relatively free of lions and of almost everything else that kills people before they can reproduce.

I know other people have blogs where they explain things and readers bask in their wisdom, but having a blog where I say how confused I am about stuff and readers explain it to me has always worked for me before and I have faith it will continue to do so.

HBD Chick has already addressed their response from pretty much the same perspective I will, but just to make it official…

I originally complained that Galton’s Victorian sample was more elite than the modern day samples in the study, and so could be expected to have higher IQ by selection bias alone. The authors counter that Galton broke his sample down into demographic subcategories, and each of those subcategories does better than the corresponding subcategory today – therefore, demographic differences are not responsible for the IQ difference.

HBD Chick pointed out, correctly, that the similarly-named categories hide some big differences. For example, Silverman directly compares Galton’s Victorian students (presumably college students, since we’re talking age 18 – 30) to 1940s college students. While both groups are college students, one is more elite than the other: according to National Center For Education Statistics, the percent of Americans with bachelors’ degrees rose from about 2% in 1900 to about 10% in the 1940s. If England was similar, and if college is pretty meritocratic, then the 1880 student sample was drawing form about the smartest 2% of people, and the 1940s student sample from about the top 10% of people.

The other comparison the authors brought up was Anger’s 1993 sample of “220 postal, hospital and insurance workers from three different US cities”, which they compared to Galton’s Clerical/Semi-Skilled and Semi-Professional groups and found that, even matched job for job, the Victorians did a bit better. This is good work, but I am still skeptical for two reasons.

First, in a society with a large underclass, the middle-class will be comparably elite. Suppose that we took the least productive 90% of people in 2013 America and put them in those power-generating pods from the Matrix. We still need postal workers, but now the postal workers are coming from what was previously the top 10% of the population – maybe people who would have been college professors in the old system. But the Victorians had a comparably larger underclass (small farmers and factory workers) and a comparably smaller middle class than we do. Therefore, we would expect Victorians with a standard middle class occupation to be drawn from more elite segments of the population than a comparable modern with the same occupation.

Second, Galton’s sample was still self-selected from people who came to a science museum in an elite city and paid him money! I propose that the average postman you will find buying science-related merchandise in the gift-shop of the Exploratorium in San Francisco will be smarter than the average postman drawn randomly from the distribution of postmen!

So although I appreciate the authors’ time and politeness in responding to me, I am still skeptical of their findings.

Let me move to one other point that many of my commenters protested. In the original article, I said:

All the theory here sort of checks out, except for the part where they say IQ changed 15 points in a hundred years, which is just a little bit faster than any responsible person expects evolution to progress. People critique the idea that Ashkenazi Jews could have shifted fifteen points in nine hundred years on the grounds that it’s too fast.

Many people correctly noted that it’s a lot harder to evolve a complex feature in the first place than it is to break one, and therefore my comparison was invalid. I actually considered this when writing the original post, but decided to leave it in. Here’s why.

In a normally distributed trait like IQ, over a short period like nine hundred years most of the action isn’t going to come from de novo mutations. It’s going to come from people on one side of the bell curve out-reproducing their peers on the other. This can happen just as quickly for smartening up as it can for dumbing down.

In my comments section, Konkvistador writes:

The sheer 15 point drop made me go ‘nope’ right away. If I remember Lynn right we should be losing about one IQ point per generation of our genetic potential in the West, two thirds of that due to mass immigration which wasn’t a big factor for Britain before the late 70s. The dysgenic drop can’t be more than about 3 points.

These seem like much more reasonable numbers. They also seem like a good reason not to worry about dysgenics ever again. How many more centuries do we have in which natural selection is going to be the main force shaping our genome, as opposed to genetic engineering or transfer to nonbiological life? Maybe one, if you’re really pessimistic? Losing three IQ points before that time (most of which are probably instantly regained via Flynn Effect or some kind of low-budget intelligence amplification like the racetams) isn’t anything to lose much sleep over.

Meanwhile, Betteridge’s Law of Headlines continues to warn us that “Any headline which ends in a question mark can be answered by the word no.”

On first glance, the paper looks solid. It investigates simple reaction time, a measure which is known to be correlated with g, the mysterious general intelligence which is supposedly measured (to some degree) by IQ tests. People have been experimenting with simple reaction time for over a century now, so the paper asked the relatively simple question of whether it has changed over that century. They found that it had: it had gone up, signifying a decrease in general intelligence. Their explanation was dysgenics.

People have known for a long time that high-IQ people have fewer children than low-IQ people, so it might make sense genetically to believe that each generation becomes a little dumber. This pattern has stubbornly refused to appear: instead, every generation has had significantly higher IQ than the one before, an observation called the Flynn Effect. This has been attributed to various things, including better nutrition, child-rearing, and education.

What the authors of this paper do – and it’s pretty clever – is say that the Flynn Effect is an environmental increase in IQ which has hidden a simultaneous genetic decline in IQ. They try to prove it by saying environmental and genetic factors affect IQ in different ways, and that genetic factors are more likely to affect certain features like reaction time – a pattern which is called a Jensen Effect and which is on relatively solid ground. Because they find reaction time is declining, probably people are becoming genetically stupider and the only reason we can keep having a civilization at all is because our environment is getting better – which is too bad, since our environment may have stopped doing that.

All the theory here sort of checks out, except for the part where they say IQ changed 15 points in a hundred years, which is just a little bit faster than any responsible person expects evolution to progress. People critique the idea that Ashkenazi Jews could have shifted fifteen points in nine hundred years on the grounds that it’s too fast. So let’s take a closer look at their data.

Only two of their sixteen studies come from the Victorian Era: Galton 1889 (n = 3410) and Thompson 1895 (n = 49).

Francis Galton, a brilliant Victorian scientist who was a half-cousin of Darwin, is the source of 98.5% of our Victorian reaction time data – not to mention the concept of reaction time itself, several statistical tools including correlation and standard deviation, the use of the survey in data collecting, the term “eugenics”, the entire science of meteorology, hearing tests, the first study on the power of prayer (he prayed over random fields to see if the crops there grew higher; they didn’t), fingerprinting, the scientific investigation of synaesthesia, and a horrible warning about how not to do facial hair.

Galton’s Data A Century Later, published in 1985, tells us a little about how he gained his ground-breaking reaction time statistics. He set up a laboratory in the Science Galleries of the South Kensington Museum. There he charged visitors to the museum three pence ($25 in modern currency after adjusting for inflation) to be measured by his instruments, a process he advertised as “for the use of those who desire to be accurately measured in many ways, either to obtain timely warning of remediable faults in development, or to learn their powers.” Over the course of nine years, he attracted about nine thousand curious individuals, three thousand of whose data managed to make it into the current meta-analysis.

His colleague in Victorian reaction-time measurement was Helen Thompson Woolley, an American psychologist who published a 1903 dissertation titled The Mental Traits of Sex: An Experimental Investigation of the Normal Mind in Men and Women (it was, apparently a simpler time). With an optimism bordering on the incredible, Wikipedia notes that “Before Woolley, research on sex differences was heavily influenced by conjecture and bias.”

Woolley writes of her sampling technique:

“In making a series of tests for comparative purposes, the first prerequisite is to obtain material that is really comparable. It has been shown that the simple sensory processes vary with age and with social condition. No one would question that this statement is true for the intellectual processes also. In order to make a trustworthy investigation of the variations due to sex alone, therefore, it is essential to secure as material for experimentation, individuals of both sexes who are near the sae age, who have the same social status, and who have been subjected to like training and social surroundings. Probably the nearest approach among adults to the ideal requirement is afforded by the undergraduate students of a coeducational university. For most of the the obtaining of an education has been the one serious business of life.

The individuals who furnished the basis for the present study were students of the University of Chicago. They were all juniors, seniors, or students in the first year of their graduate work. The subjects were obtained by requesting members of the classes in introductory psychology and ethics to serve.”

She found (a finding replicated by all later studies and now considered essentially proven) that women have slower reaction times than men (interestingly, this difference does not correlate with IQ) – but more relevant to the current meta-analysis, she found the same generally fast reaction times as Galton.

The modern studies, keeping with the zeitgeist of the modern age, are much less colorful. I only looked into the two largest: one Scottish, the other Australian. Here’s what the Scottish study says of its methodology:

The study was originally located in the Central Clydeside Conurbation (Figure 3), a socially heterogeneous and predominantly urban region, including Glasgow City, which is known to have generally poor health. Two-stage stratified sampling was used to select subjects. For the regional sample, local government districts were stratified by unemployment and socio-economic group data from the 1981 Census and 52 postcode sectors were systematically selected from these with a probability proportionate to their population size. The same postcode sectors were chosen for all three cohorts. The sampling frame used for individuals was Strathclyde Regional Council’s 1986 Voluntary Population Survey—an enhanced electoral register that provides details of the age and sex of all household members.3 Individuals were selected from the 52 postcode sectors within each age cohort with a systematic selection with a prescribed sampling interval from a random start.

I was getting bored by the time I made it to the Australian study, but I managed to keep my attention on it long enough to note the following sentence:

Persons selected at random from the Electoral Roll [of Canberra] were sent a letter informing them about the survey and saying that an interviewer would contact them soon to see if they wanted to participate.

Look around you. Just look around you. Have you worked out what we’re looking for yet?

That’s right. The answer is selection bias.

Back in the Victorian Age, science was done by aristocrats and gentlemen who drew their subjects from their own social groups. There were no poor people in either study, because getting poor people to participate in an experiment would require finding some poor people, who probably smelled terrible and lived in areas where there were no good restaurants.

In the Modern Age, everyone is excruciatingly Socially Aware, and studies go out of their way to look at Disadvantaged Disempowered Disprivileged Populations so their results can serve as Cutting Social Commentary.

Galton’s study population was visitors to a science museum in the posh part of London who were willing to pay him $25 to participate. Thompson’s population was University of Chicago philosophy students. The two modern studies are random selections double-checked to make sure they don’t undersample the poorest sections of the population.

So, uh, congratulations, authors of this paper! You have successfully proven that the average member of the population is dumber than wealthy science dilettantes and students at elite colleges! Go pat yourself on the back!

In case we need more rigor: according to The National Center for Education Statistics, about 2.3% of Americans went to college in 1900. In a perfect meritocracy maybe only the smartest people would go to college, but we’re not a perfect meritocracy. Would it sound about fair to say that the people in college at the time were a sample of the 20% or so of the smartest Americans?

Because the IQ of someone at the 80th percentile is 113 – that is, exactly enough to explain the 14 point IQ “drop” that Woodley et al found.

This is a little harder to do with Galton’s science museum visitors. The 1985 commentary on Galton’s data tells us:

As would be expected of a group of paying testees being measured in a museum, a sizable portion of Galton’s sample consisted of professionals, semiprofessionals, and students. However, as may be discerned in Tables 10 and 11, all socioeconomic strata were represented.

Tables 10 and 11 turn out to be a gold mine – I worried the records of exactly who took the tests would be lost, but as you might expect of someone who basically invented statistics single-handedly and then beat Darwin in a debate about evolution as an encore, Galton was very good at keeping careful data.

This site tells me that about 3% of Victorians were “professionals” of one sort or another. But about 16% of Galton’s non-student visitors identified as that group. These students themselves (Galton calls them “students and scholars”, I don’t know what the distinction is) made up 44% of the sample – because the data was limited to those 16+, I believe these were mostly college students – aka once again the top few percent of society. Unskilled laborers, who made up 75% of Victorian society, made up less than four percent of Galton’s sample!

So this discredits this meta-analysis way beyond any need for further discrediting, but since I can’t help beating a dead horse…

Let’s talk about race. We know that studies find white people usually have faster reaction times than black people – in fact, a lot of the voluminous and labyrinthine research on race and IQ hinges on this fact. We thankfully do not have to enter the minefield of trying to figure out the causes of this discrepancy (biological vs. environmental vs. social) – we can just take it as a brute fact.

What percent of Galton’s 1889 science museum visitors do you think were non-white? What percent of Thompson’s 1895 University of Chicago students? Approximately zero? Sad to say, non-white people were as likely to be exhibits in the science museums of the day as visitors, and according to no less a figure than W.E.B. DuBois in 1900 there were only 2600 living black Americans who had graduated college.

I looked them up some stats on the sample areas for the modern studies – 6% of Glasgow is non-white, and about 12% of Canberra. So aside from selection bias affecting intelligence which affects reaction time, we have selection bias affecting race which affects reaction time.

May I just say how annoyed I am that I have to remind reactionary eugenicist IQ researchers, of all people, to pay attention to race? YOU HAD ONE JOB!

Finally, there’s significant IQ differences within populations of the same race and country simply due to migration effects. An analysis of IQs across Great Britain finds that the highest scores are in London (102) and the lowest in Scotland (97). Almost all this meta-analysis’ Victorian data came from London (Galton’s museum in Kensington) and the largest source of modern data (making up about half of the whole, and being unusually high in reaction time) came from Scotland (and Glasgow isn’t even the nice part of Scotland). The 5 point London – Scotland difference explains over a third of the “difference between Victorians and moderns” found in this study.

So in conclusion, this study ignores race, ignores regional variations, but most importantly IGNORES THAT ALL ITS VICTORIAN STUDIES WERE SAMPLING FROM THE SMARTEST 20% OR SO OF THE POPULATION AND THEY GOT EXACTLY THE NUMBERS YOU WOULD EXPECT IF YOU DID THAT.

There is some really excellent IQ research out there that everyone should be reading, but this is not it. Please please please don’t cite this study as evidence for dysgenics or the decline of civilization.