The whistleblowing arms race »« Film review: War on Whistleblowers

The puzzle of rising IQ scores

Nearly three decades ago, work by James R. Flynn revealed that average IQ scores in developed countries were rising at a stunning pace, of the order of 0.3 points per year or more. Later work showed similar explosive gains in developing countries and that the rise (now dubbed ‘the Flynn effect’) is still continuing. How did he find this? Recall that although IQ tests themselves have changed over time, each revision requires the IQ scores to be normed to always have an average value of 100. So does someone who scores 100 today have the same IQ of someone who scored 100 say fifty years ago, since they took different tests? Flynn found that if you give people today old IQ tests, their scores rose steadily the older the tests, suggesting that IQ tests have got harder over time. (James R. Flynn, Are We Getting Smarter?, 2012)

It seems highly implausible that biologically-based raw intelligence could explain this rise. Evolution just does not work that fast. So what could be the explanation? One possibility is that over time education has become more widespread. More people are going to school for longer periods and this may result in more of them learning the kinds of things that help them do well on IQ tests. But a more important reason may be that over time, people have increased their ability to think in terms of abstract quantities and relationships rather than in concrete forms, not just because they have more schooling but because the emphasis of the education they receive has changed.

For example, if you ask the question of what dogs and rabbits have in common, people nowadays would answer that they are both mammals, rather than that they both get involved in chases, with dogs pursuing rabbits. The former answer would be deemed correct while the latter would be marked wrong. But the practice of imputing abstract properties to concrete objects and then using those as points of analysis was not that widespread in the past and is now the product of an increasingly science-based education. Flynn says that while we now commonly use such ‘scientific spectacles’ in answering questions, in earlier times and in other communities, ‘utilitarian spectacles’ were more used.

He gives the account of researcher A. R. Luria who published transcripts of interviews with rural Russian people in the 1920s that illustrate the difference.

Q: What do a fish and a crow have in common?
A: A fish – it lives in water. A crow flies. If the fish just lies on top of the water, the crow could peck at it. A crow can eat a fish but a fish can’t eat a crow.
Q: Could you use one word for them both?
A: If you call them “animals,” that wouldn’t be right. A fish isn’t an animal and a crow isn’t either. A crow can eat a fish but a fish can’t eat a bird. A person can eat a fish but not a crow.

Commenter jpmeyer gave another example (maybe also from Luria) in response to an earlier post that shows how the ‘correct’ answer depends on whether one is wearing scientific or utilitarian spectacles.

I wish I could remember where I read this, but I remember reading that one of the reasons why people in places like Sub-Saharan Africa get horrible scores on IQ tests is because the test will have a question like “Which of these animals is not like the others: Cow, Chicken, Pig, and Dog”. The answer you’re supposed to give is “chicken” because it’s a bird and the other three are mammals, while they would answer “dog” because the other three are animals that they farm for meat.

Here is another interview from Luria that demonstrates how logical thinking has also become more abstract over time, with increased ability to think in terms of hypotheticals.

Q: There are no camels in Germany; the city of B is in Germany; are there camels there or not?
A: I don’t know, I have never seen German villages. If B is a large city, there should be camels there.
Q: But what if there aren’t any in all of Germany?
A: If B is a village, there is probably no room for camels.

Because syllogistic reasoning is almost second nature to us today, the rural Russian seems particularly obtuse in not being able to answer ‘correctly’. But that is because we are taught to wear ‘scientific spectacles’ from an early age. Hence we can easily separate logic from concrete facts and recognize that ‘camels’ and ‘Germany’ and ‘city B’ are just symbols that have no significance and could be replaced with other symbols without loss of meaning. But this way of thinking was not so universal even a century ago. Modern IQ tests like the Ravens and Wechler use symbolic reasoning largely or even exclusively and so favor those who are familiar with manipulating them.

Flynn describes the frustration that he and his brother had in trying to change their father’s reactionary views on race in order to get him to oppose discrimination. They tried appealing to his sense of empathy by asking him, “But what if your skin turned black?” His father (who was born in 1885) was incredulous that they would make such an argument, saying “That is the dumbest thing you have ever said – whom do you know whose skin has ever turned black?” Flynn says that he never encounters any modern racist who would argue that way. We all know now to take seriously the premises of hypotheticals, however absurd, as a basis for discussion.

Another reason for an increased ability to think in the abstract using symbols and hypotheticals is that our lives in general have become more complex and require more symbolic and abstract thinking even in everyday lives. Flynn points out that even when it comes to popular culture like TV shows, nowadays they routinely feature fairly complex multiple interweaving story lines even in sit-coms, unlike the simpler and more linear ones that were the norm even as late as the 1970s. Our work and leisure activities and the technology we use today all require us to manipulate symbols and use abstract reasoning to a far greater degree than people in the past or even people now who live in different cultures.

While all these can explain why IQ scores can rise rapidly over a short time period without a corresponding increase in intelligence, they can also explain why IQ scores can differ between nations and cultures and linguistic groups without corresponding differences in intelligence. Hence we should be highly skeptical of policy decisions based on assertions about the immutability of IQ scores or the differences in average scores between groups.


  1. Brian Wesley says

    The odd thing is, I have found that in many, many internet arguments, lots of people don’t understand a hypothetical analogy. Here’s a real example from an argument about the Boy Scouts excluding gays, and when public schools used to charter BSA units:

    Me: “… it’s the same harm as if a public school ran a “no Jews” club. Now, if you’d like to argue that having “no Jews” clubs run by public schools don’t harm Jewish kids and don’t foment hatred against Jewish kids, go right ahead. You pretty much have to take that position to stay aligned with your stance regarding atheists.”

    Idiot: “We welcome Jewish kids. We welcome all people of all faiths. We just believe that believers are people we want in our organization.”

  2. Dunc says

    It seems highly implausible that biologically-based raw intelligence could explain this rise.

    I’m not sure that’s entirely correct… Unless by “biologically-based” you mean “100% genetically determined”. Take height for example – height is very definitely a biological characteristic, which can be empirically measured with no ambiguity, yet in certain populations, it shows remarkable rates of change. This is because of the fact that, whilst an individual’s potential height is (largely) genetically pre-determined, the actual height they achieve is fairly strongly influenced by other factors, most notably childhood nutrition. So it may be possible that at least some proportion of the apparent increases in intelligence might be explained by changes in non-genetic biological factors, such as nutrition… However, this does not invalidate any of the other very important points you make here. I just think it’s important to remember that there’s more to biology than genetics, and that complex traits such as intelligence usually involve multiple interacting factors.

  3. says

    I don’t understand what the big deal about this finding is, it’s easy to explain thus: IQ doesn’t measure intelligence.

    Therefore there’s no question that IQ can keep going up and up at whatever rate it does, because it’s apparently not actually measuring anything about underlying human capabilities. It’s probably measuring something about learned behaviors, but we can’t even know that. The only thing we can know about IQ is that measures a person’s ability at taking IQ tests.

    Social “science” – pfffffffff!

  4. smrnda says

    As a person with a background in mathematics and logic, the idea that these are things that are learned seems pretty obvious – when I was an undergrad, I took a cognitive psychology course which had a unit about testing ‘intelligence.’ What I found was that many of the questions were just straightforward applications of things that I had learned and that, if a person had been shown the right stuff, they’d answer the questions correctly. I recall one tricky question that almost nobody got but which required what was a pretty basic technique in graph theory. To me, this makes these alleged ‘abstract reasoning’ tests just a different sort of specific knowledge test.

    I’m also going to second that though evolution might not work that fast, improved health, prenatal care, and earlier schooling probably has a real impact on the physical brain itself, the same way that lead exposure is correlated with lower IQ scores.

  5. says

    @Marcus Ranum #3 – There is actually a more basic issue: what is intelligence? The term is used differently in different fields, with a zoologist having a different understanding than an anthropologist, both of which are different from the one used by a sociologist. Are we talking about problem solving? The ability to think abstractly? Creativity? Pattern matching? Memory? Intuition? Cognitive speed? Comprehension speed? Spacial understanding? Logical processing?

    Each of the standard IQ uses a different definition of intelligence, and test different things in different amounts. This has led to the odd results where one test shows a person to be in the 90th percentile while a different test shows the same person to be in the 60th percentile. This, in turn, leads to unavoidable cultural bias: a North African Beduin, raised in a semi-nomadic lifestyle of tracking through the Sahara Desert, will probably score higher than an urban American on tests that emphasize memory and special cognition, and lower on a test that emphasizes arithmetic skills and logic.

    My theory for the perceived increase is that it reflects trends in intelligence psychometrics towards a broader, more inclusive definition of intelligence, and the use tests that either reduce cultural bias or are calibrated to different cultural backgrounds. These would explain the global increase, which includes regions where education techniques and content have not changed much in generations.

  6. sc_770d159609e0f8deaa72849e3731a29d says

    When Liam Hudson studied psychology all psychology students took an IQ test at the beginning of the course. Thirty years later Hudson re-examined these tests. He found that the only thing that correlated positively with high IQ was belief that IQ was a reliable measure of intelligence.

  7. TheBrummell says

    IQ was explained to me, possibly by my psychology-undergrad girlfriend-at-the-time that IQ was poorly named; it was more accurate to describe it as a measure of a person’s ability or potential ability at further formal education. Call it EQ, instead, for “Educatability Quotient”. Under this view, IQ or EQ corresponds to the value one should expect to get out of another year of schooling, within the education system one is already a part of.

    Setting that aside, the part about use of abstract, interchangeable symbols and hypotheticals is really interesting. Utilitarian vs. Scientific reasoning seems to be mostly a matter of understanding the rules of the test – I expect I could score about as well on an Utilitarian IQ test as on a Scientific IQ test, if I understood which set of goggles the test-marker was going to be wearing when marking it. Dogs are not farmed for meat, but are mammals, so my answer to that question would of course depend on the cultural background of the test. Knowing to apply my inference of the test-marker’s cultural background is itself a kind of intelligence exercise, I think.

    When are hypotheticals explicitely taught? Small children, like 5-8 years old? Or later? The first example I can think of would be something like a teacher instructing the class to imagine the school being much closer or much further away from the students’ homes, and then leading them through what else in their daily lives would change. I don’t have any specific memories of such classroom activities when I was young, but it’s plausible to me that either such activities were carried out, or that I would have been capable of thinking that way at the age of 7 or so. Or maybe I’m just projecting too much on my younger self.

  8. jk says

    This article assumes that it is ‘unlikely’ that the humans could get smarter.

    People are getting taller. Why is it unlikely for brains to get better but not for legs to get longer?


  1. [...] The puzzle of rising IQ scores Meet Your New Symbionts: Trillions of Viruses The s**t hits the fan – FDA, INDs, and fecal microbiota transplants The technology that links taxonomy and Star Trek (overly optimistic; read the comments) What’s the Matter With Portland? The city has been fighting fluoridation for 50 years. Will facts trump fear this month? [...]

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>