The IQ myth

by Sarah Belisamu

From its inception as a research tool, the concept of IQ has grown into a worldwide industry, that has formed the basis of recruitment and popular psychology but has sometimes led to uncomfortable and spurious generalizations about groups of people, and more recently has been the subject of DNA profiling.

We are very far away from understanding how the brain works, or the DNA governed mechanisms that produce life. We have 20000- 25000 genes, and it is known that at least 500 have an influence on intelligence, but the likely number may be a lot more than this, perhaps even the majority of genes have an impact on intelligence, as many processes are generic and function in many cells in the body including brain cells. It is estimated that a typical genome differs by 1.2 million bases from a reference genome. The number of combinations of DNA exceeds the number of people in the planet by many orders of magnitude. There are many processes behind this variation and whilst there may be more deleterious examples of variation arising from deletions of genes, fragments, or even whole chromosomes, that may lead to untimely death, poor health, reduced fertility and impact on intellectual function, there are infinitely many more, conserved, even handed variation of genes, before even considering the synergism that might exist between combinations of genetic mutations. With the dice being thrown so many times it is statistically impossible to inherit every ‘intellectually beneficial gene’ and our own genetic inheritance when compared to others cannot be deemed to be superior. In a sense we are all equally unlucky or unlucky. If a coin is flipped a million times, then you always arrive at near enough to 50% heads or 50% tails, in fact 95% of the time you get heads between 49.9% and 50.1% of the time, and 49% < or > 51% heads are practically impossible outcomes. This kind of idea of the binomial distribution arising from discrete entities i.e. coins or genes is implicit in the idea of the bell shaped curve. This would be feasible if there were a few genes that governed intelligence, but that does not seem to be the case as they would have been discovered already.

Genome wide analysis has yielded some insight but currently only accounts for 1.5% of the variation in IQ. The modelling of a trait by compiling the effect of many genes is called polygenic scoring and already it is becoming a big business, despite the lack of predictability that is currently possible. A worrying aspect is that such tests could potentially be used to screen embryos and sperm cells used for IVF and profiles fetuses for desirable traits. Nature has already developed stunning mechanisms to maintain the quality of DNA and it is naïve to think we can facilitate this through intervention when we have such an incomplete understanding of the underlying processes. With so many genes involved the idea of being advantaged melts away within the incredible number of permutations.

A remarkable trait that we possess is the universal ability for us to communicate with each other from a young age, despite language being complex and nuanced, and most of the time we get it instantly, even when there is misdirection, subtext or double entendres. We all laugh at hearing some joke that we have never heard before, perhaps there was some inventive use of language. Language is evolving all the time and we evolve with it and contribute to its evolution and yes some might have a much broader vocabulary than others but does this matter? Using language that most people do not understand impairs communication, and in many cases, it is better to embrace normality and use language that is understood by the majority. Consider, have you ever encountered an unfamiliar word and still not understood it after looking it up in a dictionary? I would venture that the answer is no, and perhaps that word is somewhat redundant and might not be worthy of its entry in a dictionary. 

Another problem with IQ test is the very narrow spectrum of intellect that is being assessed. IQ tests rely on some analytic or pattern recognition skill, but responses rely on picking an option, but we do not spend our lives evaluating a limited range of options for which we have complete information, or ticking boxes, nor is the latter activity valued much except perhaps during elections. Our most developed, varied, interesting and profitable intellectual function is our behavior. We don’t spend time ticking boxes, we interact, we influence and some of us have developed unique skills, from kicking a football to dancing, acting, oration. These are not physical acts but incredible intellectual feat too. When a footballer calculates the velocity and angle of a strike instantaneously that is an intellectual task. No one is the best at more than one thing. Messi might play chess and tennis, but he isn’t better than Carlsen or Djokovic, and for anyone that has developed a successful business have they not defied the odds, gained some unique understanding of their product, their clients, marketing and promotions and communication skills. None of these feats could ever be accomplished without hard work and these are all intellectual feats that are valued. But these aspects of behavior are very difficult to measure. To measure global intelligence, would necessitate measuring all these abilities and adding up all these scores with some arbitrary weighting, but then again what would be the point of such an exercise? What is the meaning of the overall score if you lose sight of the incredible feats, if you didn’t see the goal hit the net, Joan Rivers’ mischievous punchline delivered with a cackle, or layers of paint on a Van Gogh canvas? If you tried telling any of these individuals that they achieved eminence in their field without hard work, then they would scoff at the suggestion. The greatest achievements are borne out of love and obsession.

IQ certainly has a significant heritable component but what is most peculiar is the dogmatic insistence that it is somehow fixed, or it can’t be changed much like height once we reach adulthood. Any intellectual skill can be improved or practiced. People can and do change but only if they want to. In test-retest studies of IQ tests. though there is good correlation, there can be huge changes to individual scores. If IQ was largely genetically determined, then this would not be possible for anyone. On average people stay the same, it is their own choice, they adapt to the demands of their own environment and psychological needs and if that means spending more time on a football field and less with their nose buried in a textbook, and this led to them becoming successful, then we cannot deny that they have behaved in an intelligent manner. We are not measuring intelligence directly just an application of intelligence. Intelligence is the ability to control the environment both internal and external to achieve, happiness, health survival and success and if certain intellectual skills do not facilitate this then why retain them? These skills like genes will fall by the wayside if they do not promote survival. Likewise with the passage of time, new technologies, social conventions and trends, certain skills become obsolete whilst others new skills come to be valued. The brain’s hardware is indeed very powerful and there is an immense amount that can be achieved. If you take a computing analogy, then a software developer who berates their hardware probably isn't a very good one. There is still an incredible amount that can be accomplished within hardware limitations and if a computer has a faster processor and runs faster than another, is it more intelligent if it runs the same software? Or is it just an automaton that travels the same route at a faster pace? The intelligence is in the software and the program is something that we develop, and for those things that spark our interest, that program is very complex. We learn to understand complex ides by breaking them into a series of smaller steps and we only need to do this once, and then we can create a shorthand for that understanding which we can use as a building block for other ideas. We develop a sophisticated language for our ideas. In many ways this is the real development of vocabulary rather than remembering some word that no one else uses. Sure, there may be genetic influences behind how we develop our understanding of the world, but it can’t be denied that this a willful process. Environment is often regarded as an external factor in the development of intelligence and whilst there may be genetic process behind how we chose our ‘environment’ as well as the actual early life environment we are provided by our families, the biggest influence on the environment are our own choices.

Sometimes we gain insights through moments of inspiration, where we learn some new perspective and remodel a problem. John Lennon said, ‘there’s nothing you can say that has hasn’t already been said’ and he is right most of the time, as usually these insights prove only to be for our own personal benefit. Very occasionally they may prove ground-breaking, but these unpredictable eureka moments are usually borne out of hours of prior cognition where we are solving some problem in our own mind. The satisfaction that we have solved some taxing problem might be enough, but for many people that isn’t enough. To be successful means solving someone else’s problem simultaneously, if you want others to recognize you work rather than it merely existing on some intellectual plane. We would not have Newton’s Laws if he could not persuade others of the merits of his work. Who would have stored, archived, reproduced his writing and propagated his work? Why are we persuaded to read his work today? It must be the result of more than the effort of the work itself. Perhaps Newton’s Laws were previously discovered by someone else who never bothered to write them down, was not able to persuade someone to publish their book, archive their work in a secure location, value their work and publicize it.

IQ may be seen as liberating, as it may inspire people from disadvantaged backgrounds by challenging the rigid structures of class, stock and caste. This may be a good thing, but it is still only replacing one form of birth right with another. Nothing can be less inspiring than judging people and telling them that the die has been cast already, that they can’t change, or they can’t achieve, but it is only slightly better than telling someone that they have got it made, that they don’t have to try and they are destined for greatness.

So IQ has a normal distribution, and that means it’s a fundamentally a biological process right? - Well not really. The raw scores from an IQ test generally do not have a normal distribution. The test score must be converted into an IQ score and then the IQ has a normal distribution, but only because IQ has been pre-defined as having a normal distribution. This is an artificial process of calibration, essentially the raw score is converted into a rank depending on the proportion of people who achieve a particular score, which is converting a continuous variable into an ordinal variable. The distribution of ordinal variables has no meaning. For example, instead of measuring the time that someone takes to complete a marathon you could generate an RQ score (running quotient) by assigning points bases on the proportion who were in the top 2.5%, 20% etc. This process could be carried out regardless of the actual distribution of times (which is not normally distributed but has a positive skew towards longer times in this example). If you arbitrarily select the right percentages for the right points to match a normal distribution, then a normal distribution could be artificially assigned to any ranking scale. Some IQ tests do not need to be adjusted very much but others are more skewed. In the picture above the frequency distribution as a percentage for the raw scores on a Weschler Intelligence Scale for children, for Full Scale IQ (blue) and verbal IQ (orange) is plotted. As you can see it is very far from a normal distribution and both scales have a double hump like a Bactrian camel. The example of a marathon race might seem ridiculous. The most ridiculous aspect is that all the raw data for the times are lost, and the curve here has no meaning; there was always going to be someone who came first and someone who came last, and 10% of people were in the top 10 percent – we knew that before the race was run. A number can be applied to anything, but this does not necessitate that it carries any meaning. Height has a true normal distribution; no such transformation is required to produce a height quotient. What is the meaning of IQ having a normal distribution if the results of the tests that measure intelligence do not have a normal distribution? And does a normal distribution really imply that there is some fundamental biological process; not really- there is no fundamental process behind tossing a coin. Challenging the normal distribution may be regarded as a moot point, but the bell curve provides a powerful icon to promote the concept of IQ, and the idea of ranking and judging people. The idea would not have been as appealing if the distribution looked more like a Bactrian camel than a bell-shaped curve.

I am not suggesting that psychological test have no merit, just that the idea that you can sum up the most complex thing that we know by one number is absurd. I don’t think the idea of ranking people is a healthy idea or that intellectual skills can’t be gained or lost to a large degree. Intelligence tests measure a tiny proportion of intellectual abilities and should be viewed more within the wider context of behavioral attributes where there is no sense of ranking just a palette of personality attribute or tendencies. Life is game enough. It’s high time to take this minigame with a hefty pinch of salt and let no one be judged.