IQ is a dismal measure of intelligence

Jump to navigation Jump to search

This article by Dr Piotr Wozniak is part of SuperMemo Guru series on memory, learning, creativity, and problem solving.

Problems with IQ

The concept of IQ seems to be increasingly harmful. It is time to ditch the popular culture of IQ tests. IQ is an excellent research tool, however, it has awful psychological and sociological implications.

Here are my main concerns regarding IQ:

  • IQ is remotely relevant to actual intelligence
  • IQ is popularly perceived as determined by the genes and unchangeable
  • IQ tests have a negative effect on test takers
  • IQ is often used in "scientific" racism to promote pseudoscientific social theories

The theoretical inadequacy of the IQ derives from the fact that the brain is a concept network whose performance is always domain-specific. Performance of concept networks is only marginally better if we improve the speed of processing, and that speed is largely constant in a healthy population.

Similarly to proponents of schooling who see the brain as a hard disk to which we need to copy the curriculum, proponents of IQ see the brain as a microprocessor whose speed can be measured with a benchmark. Those mechanistic misinterpretations fail to notice that the brain is a concept network

Intelligence cannot be separated from knowledge

A vast majority of people and researchers attempt to separate intelligence from knowledge. It is a futile effort. Raymond Cattell used the term problem solving when speaking of fluid intelligence. He wanted it to be seen as separate from knowledge-based crystallized intelligence. However, all problem solving requires knowledge. IQ tests attempt to dig down to the lowest level of skill to extract the illusion of touching all rudimentary skills and only rudimentary skills. In reality, they miss a lot and include too much.

We can compare the speed of two processors between two computers. However, the true intelligence of computers is in their software. The speed only determines the waiting time to get the answers. Moreover, human brains are far closer in the distribution of their sheer low-level performance than microprocessors within a single class of a single generation. We all have similar neurons with similar properties. Genetic handicaps affect a minority. The wiring of the brain, esp. as determined in childhood can provide a serious handicap too. It is not because some people are more intelligent. It is because of the injurious environment we push our kids into. In addition to various forms of abuse, we undermine kid development by the limits on freedom at daycare, kindergarten, preschool, and early schooling. True human intelligence has little to do with the speed of processing, working memory, attention, etc. On the other hand, basic micro-problem solving skills are a derivative of macro-problem solving skills. A proficient adept of incremental reading will do well on reading tests, because of the fact that incremental reading, which helps build intelligence, will also provide a great deal of training in reading. This provides an illusion that good reading skills are a good measure of intelligence. In reality, even severely handicapped children can master reading.

IQ is no measure of intelligence

Intelligence is the ability to solve problems. When we speak of intelligent civilizations, we mean civilizations that have managed to accomplished great feats in science and engineering, or perhaps culture. When we speak of intelligent people, we should rely on their intellectual accomplishments, not on flashes of genius (e.g. in memorizing a deck of cards). When we speak of intelligent children, we should focus on appreciating their progress.

Basic intellectual skills measured by IQ correlate with intelligence only because these are the tools that are often involved in skills that contribute to intelligence. At the same time, problem solving relies on creativity, which may turn out to be an obstacle in test taking. Daydreaming is an anathema of a testing procedure with a ticking clock. At the same time, daydreaming contributed to many of human's highest achievements in science, engineering, arts, and beyond.

I cannot solve a Rubik's cube, and I refuse the claim that this basic problem solving skill be a measure of my imperfect intelligence. Even more so, I refuse the idea that Rubic's cube training is a necessary ingredient of my further progress in my involvement in science. IQ tests might be as well replaced with tests of how fast people can count nickels. The only problem is that this skill is too trainable to qualify. Nobody can beat a shop assistant in nickel counting. We would thus need a test with a set of skills of which nickel counting would only be one. I am awfully bad at nickels because I can never focus well on activities that I consider a waste of time. In addition, not many people in Europe actually know how much is a nickel or how many ounces in the pound.

True human intelligence is based on a vast reservoir of abstract knowledge applicable in problem solving
Personal anecdote. Why use anecdotes?
I did try an IQ test in the past a few times. Results were all over the place depending on the time of day I took the test. I will not disclose the results, because I do not want this text to be skewed by a number I am criticizing. If the number was low, skeptics would say I am not smart enough to appreciate the value of IQ or that my claim is just sour grapes. Jordan Peterson would stop reading. If the number was high, skeptics would say that only high IQ people can solve problems such as how to learn optimally (see: History of spaced repetition on my personal claim to fame). The truth about inventing spaced repetition is that it did not require high intelligence. It was just a coincidence of a high learn drive (which may be more valuable than high IQ), and a bit of mathematical intuition at the time when I was deep into learning mathematics for my computer science entry exams. Considering my tendency to being a bit boastful and arrogant, I bet you may tend to think that my IQ test went bad because I would rather boast of it? However, a narcissistic individual would not really mind tossing in an outrageously high number. Nobody would test to falsify it. My IQ can actually be found at SuperMemo Guru, but I won't list it in this text. It should not really matter

IQ is unreliable

The most absurd part of IQ tests is that they involve speed. They do not test who is a better problem solver (e.g. to figure out the laws of gravity), but who is a faster problem solver (e.g. to multiply numbers in memory). If speed is involved, we would need to consider other variables as the state of health, quality of prior sleep, and the circadian phase. Depending on the time of the day, intellectual performance may improve wildly in a number of measures. Circadian phase may translate to 50 IQ points in people who adhere well to natural creativity cycle. Why don't IQ tests consider that? Why don't experts speak about it? It comes from just a poor understanding of the power of the circadian cycle as well as from the old soup problem that is as relevant to bad schooling as it is to bad sleep. A great number of people live lives on average sleep quality only because they do not tangibly see the exponential impact of good sleep on intellectual performance. The same individual taking the same test at 10 am, and at 3 am may shows up as highly gifted or just enough to be a "farm laborer" (I borrow the ugly language from the lesser IQ theorists). For similar reasons, we do have alcohol tests for drivers, but no obligatory alertness tests, even though sleepiness competes with alcohol for the number of deaths on the road.

If IQ tests do not consider the circadian factor, they are a very poor measure of intellectual agility. We would not accept an IQ test result from a guy who has just downed six cans of beer, and yet we do it frequently from people who are in the condition of sleep deprivation.

If speed is involved in the test, a major handicap may hit creative individuals. Creativity is essential in problem solving. Creativity is part of human intelligence. And yet, a creative individuals, in their excitable brain states, may fall victim to inattention, or OCD-like iterations. Many smart individuals will recall moments from their high stake tests and the following mental loop: "the clocks is ticking, there is less the less time to solve the problem, I would love to solve it fast but ... the clock is ticking, I would ... but ... the clock .. but ... " (ad infinitum). That's a futile iteration on concept map activations, better known as the "school test panic". Creative individuals, in high excitability states, will be prime victims of futile loops in concept map activations. Their prime weapon of intelligence, will also be a prime handicap in an IQ test.

Reliance on speed, and the failure to account for the brain's cognitive state, disqualify IQ tests as a reliable measure of intelligence

Intelligence can be boosted

It is true that it is hard to improve one's IQ scores. Good sleep, good circadian timing, and lots of practice in taking IQ tests can instantly add several points. As of then, improvement is possible via training, but is limited by the same difficulties as the speed of solving the Rubik's cube, or the speed of the 100 meters dash. Major gains can be achieved by boosting one's mnemonic skills or mental calculation skills. However, those gains come with hard labor that few would be ready to undertake. Most of all, those gains do not translate to major advantages in life unless targeted at specific applications (e.g. mnemonic techniques can be a blessing in anatomy exams in a medical school).

Basic intellectual skills do not improve much with training unless a major investment of time is committed. This type of investment will not be very useful for intelligence. Solving a Rubik's cube in record time will not have much effect on one's chances of finding a cure for cancer. Instead, smart and efficient study of the science of cancer is the right way to go. The obsession with an IQ score is a mis-investment.

Instead, we need to realize that what society really needs is excellent problem solvers in all fields of human interest. Problem solving skills cannot easily be measured, however, we can easily see the results when they come. Problems solved are a badge of honor and everyone can aspire at collecting many such badges. True intelligence is a target of aspirations for everyone. Everyone can make the first step today, and aspire to genius accomplishments in a decade or two.

If I practice memorizing card decks to boost my mnemonic skills, I do marginally add to my working memory power. However, when I read about the Parkinson's disease and see how it fits the concept of the war of the networks, I tangibly improve the understanding of the human brain. This makes me a better problem solver in the area of neurophysiology. A few minutes of reading can advance my intelligence. In the meantime, I might be forgetting many lesser facts, but even this forgetting can chisel out the essence of what I know about the universe. Given good health and diligent effort, my intelligence keeps increasing. It helps me see how useless the concept of IQ is in measuring who I am in intellectual terms. The focus on the IQ is comparable to a singular focus on the power-to-weight ratio (hp/kg) when purchasing a family car.

The focus on the IQ is a major distraction. We should focus on high achievement based on high intelligence

The psychological harm of IQ

If IQ scores are hard to boost, individuals who score poorly, or not up to their own expectations, are left with the sadness of lesser achievement. This has awful psychological implications. For that psychological impact alone, IQ tests should be banned from schools. Once kids are labelled with imperfect IQ scores, they can be scarred for years to come. Scott Barry Kaufman explains that on his own example (see the video).

Teachers do respond differently to kids with different IQ labels even if those labels are experimentally falsified. This can lead to a self-fulfilling prophecy in which a falsely boosted IQ label has a positive impact on the actual IQ. The concept of IQ promotes inaction. For some, it is a reason to boast (and stay lazy). For most kids, it is a reason to be depressed.

Personal anecdote. Why use anecdotes?
In 1975, aged 13, I was thrown away from my school with a label of an incorrigible hooligan. I was admitted to a new school. The new school was located in a newly built district largely free from crime and other social pathologies. Miraculously, my school reports did not follow me and I somehow received a label of a good student. For a while, I loved my new role and it became a self-fulfilling prophecy. It culminated in a gold medal in Chemistry Olympics that made my whole school proud. In my case, the transformation took just a few weeks. All it took was a change of the label
For the same kid, a minor change in the label may result in a major change in outcomes

IQ myth

A good metaphor for the futility of measuring intelligence with IQ would be an attempt to compile Athletic Quotient (AQ). We might provide a set of tests on athletic performance, look for overall winner among athletes in the designed battery of tests, and then see which subsets of tests correlate bests with the winning performance in order to provide a perfect measure of athleticism. That measure might correlate with the quality of muscle fibers, reaction time, cardiac efficiency, metabolic efficiency, VO2max, etc. Who do you think would come out on top in the world of sports in the area of such designed AQ? Would it be Anthony Joshua who would pulverize Eliud Kipchoge with a single anesthetic punch, or would that be the Kenyan who would prove stronger by leaving the boxer huffing and puffing on the first kilometers of their long distance race for survival? It becomes instantly obvious that we would need a different quotient for marathoners (dominated by East Africa), and for sprinters (dominated by West Africa).

Incidentally, we already have Athletic Intelligence Quotient (AIQ) too. It is just one of those many intelligences that are most applicable in a given field. Analogously we might have surgeon's IQ, pilot's IQ, therapist's IQ, politician's IQ, and so on. That would be a better measure of suitability for pilots or surgeons over the plain all-encompassing IQ.

Cross-fit games might serve as an measure of the AQ, but then the selection of the disciplines will determine the winner. I always thought that I would do great in triathlon, if I did not have to bother with the biking part. Too bad aquathlon is not popular enough to find any events in my area. AQ is as useful as IQ or EQ. These are all interesting measures that might be used in research, but they are all overemphasized in optimizations!

Why do we rely on speed, or short-term memory, or vocabulary when measuring intelligence? The main reason is that those qualities tend to improve with hard work in the areas covered by school curriculum, and those are in turn defined by jobs that we see as most needed or appreciated in society of yesterday. Those who do not improve much at school are probably not interested in working hard, or are prevented from working hard for a reason (e.g. a disability).

In true intelligence, we need healthy creativity, healthy learn drive, and rich abstract knowledge that drives the ability to solve problems. Each domain will have its own set of criteria that will put a premium on working memory, reasoning, pattern recognition, or other low-level skills.

Instead of maximizing IQ of populations, or maximizing test scores of individuals, we should optimize for the best conditions for growth. In a good garden, green quality emerges naturally.

Child's IQ

The whole field of child IQ is riddled with bad science. The concept of mental age should be banished (unless defined in a fruitful way for the sake of research). If you consider the precocity paradox, the whole idea of comparing a kid to other kids of his age is harmful for both those with high IQ and for those with low IQ. Child prodigies receive accolades that build up expectations that are hard to meet. No wonder, we say "early to ripe, early to rot". The whole cohort of little girls come to front in praise and glory only to see the pack of rowdy boys catch up and even come to the lead. This can be pretty disheartening if we see life as a zero sum game based on numeric labels. On the other hand, those branded with low IQ may find it hard to recover. Kids that do not meet "age appropriate" standards get shunted from doctor to doctor for speech therapy, dyslexia, dyscalculia, ADHD, behavioral disorders, or even get put on the autistic spectrum for no good reason. All that while the best remedy has the form of just a bit of patience. Instead of classifying kids with a number that is hard to change, we should just provide them with rich environments to explore on their own. Healthy brains are excellent at adaptation. Handicapped brains are no less excellent at adaptation unless the handicap specifically targets the learning ability to a significant degree. Kids should get a chance to do the best of their potential. They should be as much free to learn as they should be free from pressure, incl. the institutional pressure for therapy (exerted by schools or hyper-anxious parents).

When a kid advances through percentile ranks in vocabulary counts as in the graph below, it says little of the advances of his IQ, or of the quality of his training procedures. The graph is just a reflection of normal development with a peak of cortical thickness delayed in time. Such a shift is often a sign of giftedness. The same kid might magically advance on age-corrected IQ charts despite having no special advantages in his environment as compared with the peers:

Figure: Speech delays often spur parents or educators to seek speech therapy. However, it is not unusual for a seeming disability to self-correct. A laggard may grow well ahead of its peers as in the presented example. Similar metrics may be used to reassure parents, and prevent frivolous push for therapy. Uneven or slow development underlies the precocity paradox

Schools improve IQ

There is a strong correlation between IQ and the performance at school. On one hand, smart individuals usually do well on IQ test, and they also know how to handle school to achieve their goals. However, the causation also goes the other way. There are two skillsets drilled at school helpful in good IQ scores. Simple calculations and simple reading are a staple of school food. Even most inveterate crammers will not escape a significant improvement in their counting and reading skills. Those micro-skills are also used richly in IQ tests. High proficiency in those simple operations often comes at the cost of high drudgery: doing a lot of similar boring repetitive operations at school. This in turns provides for one more characteristic needed for IQ tests: tolerance of pointless exercise. One of the worst things we learn at school is that learning must be done even if it is pointless. Tests must be completed even if they serve no higher purpose beyond yielding a number score. This leads to a harmful metacognitive skill: the ability to suppress a creative thought for the sake of the robotic execution of a prescribed algorithm.

For example, see this problem to solve: "Johnny has 3 cookies in a jar. Tommy has 2 cookies less than Johnny. Ronnie has half the cookies that Jonny and Tommy have ...". If you do not have a repulsive reaction to a test that niffs of school boredom, then you must have bagged many hours of patient bench grind at school. The future of mankind lays in great problems to solve. Great brains crave great problems with unadulterated pleasure. An appetite for "cookies in the jar" has nothing to do with greatness.

When in a creative zone, my focus on the job at hand can verge on mindlessness. On my best days, I may rewind my favorite Zakaria GPS a dozen of times until I have to give up. I have ADHD and I love it. It does not mean I cannot snap out from that mode in need. The stress of the test is one of the methods, except a creative thought may be worth a million and should always be given a chance.

IQ bloopers

In kids tested for IQ, we can encounter surprising obstacles. We may have a kid who breezes through all tests and is suddenly penalized for not understanding the word "button" in reference to clothing. For a modern kid, buttons are everywhere in electronic devices or in computer games, however, if a mom chooses clothing with zippers, the kid may trip on a button. Many kids have never heard of a postcard, and would not know how to use a rotary dial in a phone. There are preschoolers who do not know the days of the week or the sequence of the months just because they have never needed those. In my own on-line IQ test years ago, I was penalized for not knowing how many ounces go into a pound. My Advanced English collection told me that an ounce is just about 30g, and that 1 pound is just about 0.46 kg. I made a wrong guess: 1 pound is 15 ounces. Anyone who uses ounces in her daily shopping would instantly know the correct answer of 16 and beat me on time while possibly having no idea that a kilogram is a neat 1000 grams, let along expressing ounces or pounds in grams. American IQ tests are jam-packed with terms alien to other parts of the world accustomed to the metric system. It is hard to escape culture in IQ test. Kids in Somalia would need their own test that might involve skills related to their own childhood experience. Those who mock African test scores know nothing of African culture.

In one of on-line IQ tests I found the following question: "A marathon is 26.2 miles. If Sally ran a marathon in 3 hours and 3 minutes, what was her approximate pace?". The choice answers were 7, 8, 9, and 10 minutes.

This question is pretty tough for an unprepared mind facing a racing clock. It may also be hard for a wrongly prepared mind.

The first hurdle is the need to overcome the problem of the word "pace". For a well schooled kid, the meaning of the word may not matter. It is almost instinctive to divide 183 minutes by 26.2. It is just a matter of some mental mathematics.

A marathoner may instantly recognize the word "pace", but if he comes from anywhere outside the US, all answers instantly look wrong! I know I will never hit 183 minutes, but I can easily go 5 min/km. In a test, I might mumble to myself: "All answers are wrong, and my brain is in the state of panic. My IQ will turn out bad". It takes precious milliseconds to realize that in America, pace is measured in minutes per mile, not in kilometers per mile. As for the state of panic, IQ might be actually measuring the ability to recover. Good test taking strategy, known from school, is to skip questions that are tough and return later if time permits. This option may be disabled in on-line tests.

Being a marathoner is also useful in quick mathematical conversions. Those who run without electronic assistance know that all it takes is to see how many times 42 fits in 183. In America, instead of 42, we check how many times 26 fits in 183 to see that the answer 7 is the only plausible.

Last but not least, this question reminds me how much depends on the state of the brain. A person tested in conditions of sleep deprivation may score 10-30 points below her actual IQ level. The same person with a fever, tested past their normal bed time may hit an estimated Best IQ minus 85 difference (i.e. 60 points for a highly gifted individual). I multiply my distance passed by 5 pretty often to see if I hit my target pace in marathon. However, this effort becomes increasingly difficult as kilometers pass by due to exhaustion. More often than not, I can loop for ever restarting the computation when the hypoxic brain refuses to cooperate. Does my poor math indicate low intelligence then?

In that single question we can see the dependence of IQ on habits, culture, strategy, and the brain state. A conscientious student may need 6-8 seconds for this question. A marathoner from America might perhaps go down to 3 sec. Both will hover near maximum IQ measurable. A highly intelligent European marathoner accustomed to using calculators, may fail entirely and receive bottom IQ valuations for that test item. As a result, people with the identical level of intelligence, coming from different walks of life, or different cultures, in the same test, in the worst case, may obtain polar opposite results.

New forms of intelligence

Today's kids raised in a digital world would easily prove Carr's claim that "Google makes you stoopid", or Spitzer's "Digital dementia". Those grand claims of those brisk and well-educated minds do not stand a basic test of logic, however, they could easily be proven with IQ tests. A typical 8-year-old today will crave the news from a new world, and find YouTube as an excellent source of knowledge. For this new kind of kid, school textbooks are unbearable. The same kid will excel in many areas of intelligence that have no bearing on IQ tests. Prolific gamers show a whole range of cognitive skills that are alien to the prior generation. They will make excellent pilots, engineers, or CEOs. They will find programming computers easy. They will excel in digital art. Their cognitive advantage in music or in painting will have no bearing on IQ tests. They will have a good intuitive understanding of neural computing and artificial intelligence. They will compete well with artificial intelligence for jobs. Instead of reading books, they will soak in knowledge through the world of symbols, codes, formulas, and graphics. They will probably redefine incremental reading to their own tastes. The new kid will abhor an IQ test, and may do badly because of gaps in vocabulary, delayed reading age, poor linear thinking skills, poor attention, and/or poor patience. The new kid takes to IQ tests as he takes to school tests. Both are a dismal waste of time. Both are remnants of the old world designed by adults, while new challenges are a true attraction. The new precious form of intelligence will be penalized on the old style IQ test.

White supremacy

Nassim Nicholas Taleb puts it best: "IQ is an immoral measure that, while not working, can put people and groups in boxes for the rest of their lives" (see: Taleb's IQ is a pseudoscientific swindle).

White supremacy and "scientific" racism theories are just a sheer waste of time. The wise men of racial theories claim that intelligence is like height: you can improve it with nutrition, but you won't sore beyond 2 meters if 180 cm is your destiny. If they really need a simple metaphor to underlie understanding, they should rather compare brains to programmable computers whose adaptability gives them virtually unlimited powers.

I do not plan to dignify white supremacy logic with a discussion. IQ differences between races are a fact, but the research on the impact of environmental factors goes just a bit above the IQ of the supreme race proponents. This leads to endless and pointless discussions that serve no greater good. Those who try to pick holes in adoption studies or James Flynn's research, should paint their skin black and see how life looks in not-so-enlightened areas of America for a person of color. Like I said before, little change in labels may lead to a huge change in outcomes. People like Jared Taylor spent their lives with forlorn hopes driven by a powerful falsity vector stained with hate (see: Radicalization myth). It is really very hard to find a strong proponent of the hereditability of unchangeable IQ from people untainted with false racial theories.

As an exercise in humility, those who claim that women are less intelligent should try their game against Judith Polgar, or try to beat our Polish Marie Curie in Nobel Prize wins for science. Those who claim blacks are inferior and point to the IQ of Somalia, should get into any science debate with Neil deGrasse Tyson (I doubt he would accept the invitation). I guess, Nobel winning roster is predominantly white, and if so, perhaps it is an interesting question why. In the meantime, I am rather interested in great research and rarely have a faint idea of the skin color behind the greatest names in my field.

Due to the fact that intelligence is a direct outcome of passionate problem solving, we should ditch racial theories that only inhibit human effort. Instead, we should provide best label-free conditions for all individuals to thrive, independent of their starting point, limitations rigged from birth, and possible obstacles thrown at their feet by the less enlightened strata in the population.

Solving one's own IQ problem

Over the years I received a mountain of mail from people concerned with their memory, their learning capacity or their IQ. Invariably, that mail comes from people with pretty healthy brains. This is not mail from people who googled my mail address and spilled out their concerns. This is mail from people who read with comprehension and show a good command of my general claims in reference to memory, sleep, creativity, or intelligence. The first thing I ask for is to cool down. Not only is caring too much a self-defeating approach. In a vast majority of cases, I deal with people who speak about a self-imposed limitation. Memory and learning are trainable. Creativity and intelligence are a derivative of years of good learning. The biggest enemies of an attentive, confident, and happy mind are measurements and comparisons at school or in the workplace. When one's own brain becomes the subject of concern, the real problems fall out of focus. Even Bill Gates whose learning capacity is out of this world lists losing his mental faculties as his prime concern. We can stop this epidemic of self-concern by promoting healthy lifestyles with a great deal of good sleep and good learning.


We should ditch IQ from popular culture and from the pop science of intelligence. Instead we should focus on what really matters, i.e. intelligence understood as the problem solving capacity. We should stop bothering children with endless assessment of their skill. Some indiscriminate well-intended praise can do no harm. The fact is that the vast majority of healthy kids are equipped with a super-intelligent learning machine: the human brain (see: concept network). That machine is often injured in the first decade of life. This is how "low IQ strata" get populated. Instead of trying hard to bring all kids up to some hazy pseudoscientific standard, we should provide them all with the best options to thrive.

Intelligence is an important concept, and should best be measured by what an intellect can accomplish in a lifetime

For more texts on memory, learning, sleep, creativity, and problem solving, see Super Memory Guru