We studied 3691 people to bust IQ myths.
You’ve probably heard a lot of claims made about intelligence and IQ. Well, we wanted to check what’s actually true. So, we ran a giant study involving 3,691 people and 62 distinct intelligence related tasks. So, are the claims made by influencers, academics, and critics of IQ actually right? Does IQ actually measure anything at all or is it BS? And if it does measure something, do people with higher IQs actually achieve more? and are they happier or less happy? In a
Moment, we’ll find out the answer to these questions and others. But first, we have to address a more fundamental question. What actually is IQ? IQ is a measure of a person’s ability in whatever it is that intelligence tests have in common. For example, in our study, we had 62 distinct intelligence related tasks, including all sorts of things like memorization, spelling, math, vocabulary, mental rotations, puzzle solving, game playing, predicting patterns, and so on. Suppose that you’ve been given a random assortment of these
Tasks, and the goal is to assign a single number to your performance so that we can best predict how you’ll do on the rest of the tasks that you haven’t yet tried. IQ is precisely this number. It reflects the mathematical estimate of how you’ll perform on average across a wide range of intelligence tasks. The higher a person’s IQ, the better you’d expect them to perform if you were to give them a random intelligence related task they
Hadn’t seen yet. This doesn’t imply that IQ captures everything about a person’s intelligence. It’s just that IQ is designed to be the most accurate single number across a wide range of tasks. With that in mind, let’s jump into the claims. Claim one, IQ follows a normal distribution. You may have heard about the idea of a normal distribution, sometimes called a bell curve because it looks like the shape of a bell. It’s a distribution where most of the data
Points fall near the average with fewer and fewer as you go toward the edges. A common claim made about IQ is that it naturally follows a normal distribution with most people having an IQ close to average which is defined to be 100 and a moderate number of people having an IQ above 115 or below 85 and few having an IQ above 130 or below 70. Other people dispute this. They say that IQ only has a normal distribution because
It’s purposely transformed to have this shape. In fact, you can find long discussion threads of people on the internet arguing that the bell curve is artificially created based on the way the test is defined. So, who’s right? Well, it is true that in some data sets, they force a bell curve on the data. But we can look at what happens without that. In our US general population sample, we did indeed find that IQ has a normal distribution naturally. We did
Not define it to have a bell curve shape. This is simply what emerged from the data. Claim two, if you’re good at one intelligence task, it increases the chance you’ll be good at a different one. There’s a common sense notion a lot of people have that you’re either more of a math and science person or more of a language and arts person. You may find that you tend to do well in physics class or that you tend to do well in
English class. This would suggest that skill at math and skill at vocabulary would be negatively correlated. Being good at one makes you less likely to be good at the other. Is that what we find? No, it’s not. This image shows the correlation between performance on each pair of our 62 intelligence tasks. We can see the correlation between how well people did at task seven and task 13. In this rectangle here, we use blue to indicate a positive correlation, white to
Indicate no correlation, and red to indicate a negative correlation. Notice how almost the entire thing is blue, and there’s almost no red. That’s because doing well at one intelligence task was almost always positively correlated with doing well at other intelligence tasks. This is a strange thing to observe, but it closely matches other studies on IQ. It’s sometimes referred to as the positive manifold, the idea that scores on almost all cognitive tasks are positively correlated to each other. It
Doesn’t mean that there aren’t some people who are better at math and others who are better at vocabulary. There’s a lot of individual variability and each person will have personal strengths and weaknesses. What this does show though is that people who are better at math tend to actually be better at vocabulary, not worse on average. Claim three, IQ captures everything there is to know about a person’s intelligence. Sometimes when people discuss IQ, they make it sound like it’s
The full accounting of intelligence. End of story. But is this really true? Well, let’s take a look at the data from our study. The tall blue bar represents IQ. What this chart shows is that IQ accounts for about 45% of the variance of people’s performance across the 62 distinct intelligence tasks. 45% is a lot to account for with just a single number. But on the other hand, that means that 55% of the variability in performance is left unaccounted for by
IQ, which is even more. So what’s in this 55% of variability that IQ doesn’t account for? Well, first of all, there’s random noise. Someone might accidentally click on the wrong answer, have a bad night’s sleep before taking a test, or just get a lucky guess on one of the answers. IQ can’t account for these random factors. But there’s more than just random noise in this unexplained 55%. Another aspect not explained by IQ is that people differ in the skills that
They’ve learned through practice. If you spend a lot of time doing mental math problems, you’ll get better at mental math, and this will improve your ability at any related task, even if it doesn’t change your IQ. Additionally, people differ in their non-IQ related aptitudes. If you take two children with identical IQs, you’ll find that they are not equally good at different things. One child may have more natural aptitude at visualizing, for example, and another may have more natural aptitude at
Solving puzzles. It’s also worth taking into account that studies on IQ, including ours, typically only include intelligence related tasks that can easily be measured in a short amount of time on a computer or in a laboratory. It’s less clear if IQ would successfully capture other types of intelligence such as social intelligence or the natural intelligence of hunter gatherers. IQ captures a meaningful chunk of what we mean by intelligence, but it doesn’t capture all of it. Claim four, people
With higher IQs have greater achievements. If IQ is a measure of people’s ability at a wide range of intelligence tasks, you might think that people with higher IQs tend to have greater achievements in life. Is that true, though? Yes, we confirmed this in our study. While the effects were not always strong, we did find a positive link between higher IQ and a variety of forms of achievement. This includes education, grade point average in high school, income, and self-reported job
Performance. IQ is not necessarily relevant to all forms of achievement that people care about, but we did find that it positively correlates with a number of forms of achievement. Claim five, higher IQ people are happier. It may seem obvious that higher IQ people must be happier because as we’ve seen, people with higher IQs tend to have greater achievements, but some people actually claim that the opposite is true. That higher IQ people must be more miserable because they struggle to fit
In or because they refuse to believe pleasing falsehoods. So, what does our data actually say? Well, we found no relationship at all between IQ and happiness, neither positive nor negative. We even tried measuring happiness in multiple ways. Regardless of whether we asked in our study at this very moment, how happy or unhappy do you feel? Or if we asked a long series of questions about how satisfied they are with their life, neither method produced a link between IQ and happiness. While
There are some studies to find a link between IQ and happiness, it typically is either very small or non-existent. Claim six, IQ is just a measure of privilege. It’s certainly true that some terrible childhood experiences can impact IQ, such as severe lead poisoning and brain damage from head trauma, but some people go beyond this and say that IQ is merely a measure of a person’s privilege or social class. Is this true? In our data, the answer is no. We looked
At this in a variety of ways. First, we tallied up how many adverse childhood experiences each study participant reported, sometimes known as the person’s ACES score. This includes questions about whether they were swear at, abused, or hit as a child, among other things. We found no link between AC’s score and IQ. Additionally, we asked participants about what their level of wealth and social class was as a child. We found extremely weak relationships between each of these and
IQ. Interestingly, our data is not in line with some other studies which did find a link between IQ and averse childhood experiences and between IQ and childhood socioeconomic status. It’s unclear why our data contradicts these prior studies. IQ is not merely a measure of social privilege. If it were, we would have expected to find a strong correlation with these variables. However, it may well be meaningfully linked to a number of aspects of the childhood home environment, as there
Have been other studies that have found this. My hope is that more research on this topic will be conducted. Claim seven, people can actually estimate what their IQs are. On the one hand, you might think it’s relatively easy to figure out your own IQ because it will impact how well you do at a wide range of intelligence tasks such as tests and homework assignments in school. On the other hand, I think most of us have met someone who thinks
They are much smarter or much less smart than they really are. To test what’s true, we asked people to estimate their own IQs before we measured them. It turns out people only have a weak idea of what their IQs are. They are better at guessing their IQs than if they were completely guessing at random, but they aren’t all that much better. Interestingly, the people that overestimated their IQs the most tended to be the people with the lowest IQs.
This is compatible with the so-called Dunning Krueger effect. People higher in narcissism were also more likely to overestimate their IQ. Claim eight. Higher IQ people are more introverted. We’ve all heard the stereotype of the smart but introverted nerd who spends time solving math problems and playing on their computer instead of socializing. But are higher IQ people actually more introverted? In our data, they were, but only vary slightly. Some studies find a small negative link between extraversion and IQ, whereas
Others actually find none at all. A somewhat stronger effect we found in our data was a negative link between IQ and self-reported charisma. So maybe the stereotypes of socially awkward, intelligent people are not totally false. It would be interesting to see if other studies replicate this lack of charisma effect. Claim nine. Women and men are good at different types of intelligence tasks. You may have heard that women and men have different sorts of brains and therefore are good at
Different sorts of things. Is this really true? In our data, we did find slight differences in average performance between men and women in some intelligence tasks. For instance, we found that women outperformed tasks that require you to come up with words, such as listing as many words as you can quickly that meet certain criteria. On the other hand, we found that men outperformed women on average on tasks related to spatial visualization, such as mental rotation of objects. Sometimes
People claim that men and women differ in their processing speed, but we found no difference between men and women in processing speed tasks. Importantly, even though we found that women and men differed a bit in their average performance on some tasks, the distributions are heavily overlapping. It’s not accurate to say men are good at X and women are good at Y. It’s more accurate to say on average, men are a little bit better at one type of task
And women are a little bit better at another. Furthermore, it’s important to keep in mind that we shouldn’t jump to conclusions about the causes of differences like these. We can’t tell from a study like ours where these differences arise from. For instance, whether they’re cultural, genetic, or environmental. Today, we’ve done a speedrun through nine different claims about IQ. If you want to learn more about how your own mind works, you can take our cognitive assessment at
Clearerinking.org. You can also dig into our full report, which investigates a full 40 different claims about intelligence. If you found this video interesting, I’d also really appreciate it if you subscribe to this channel where you can learn lots more about this topic and others.