‘…a very stable genius’: What’s in an IQ score?

“…Throughout my life, my two greatest assets have been mental stability and being, like, really smart”, tweeted Donald Trump, defending his suitability for office.
26 January 2018


“…Throughout my life, my two greatest assets have been mental stability and being, like, really smart,” tweeted Donald Trump, defending his suitability for office...

This is not the first time the President declared his intelligence as one of his foremost qualifications for the position he holds - he has been known to call upon political opponents to compete with him in a very literal battle of wits: an IQ test. Just how highly President Trump would score on such a test may be a matter of some debate. But more importantly, what would the result of such a test truly mean? What, exactly, is intelligence and can it really be measured? Does it have any bearing on how much we should trust anyone with a (presumably) functional ‘Nuclear Button’?


Intelligence testing is by no means a new practice. The very first, slightly primitive, tests of intelligence can be traced back to the 1800s, with the term ‘mental testing’ coined in 1890 by the American psychologist James McKeen Cattell. However, the first person to attempt to measure a child’s mental abilities in a systematic way was French psychologist Alfred Binet. His aim was to develop a method of identifying children who were likely to struggle at school, so that classes could be streamed by teachers. Though Binet did not attempt to define intelligence, he took a common sense approach to its quantification: he judged intelligence to consist of several distinct skills, as such needing to be measured in a way that would be representative of this quality. With the help of Theodore Simon, he compiled a series of tasks which he judged to be typical of the problem-solving ability of an ‘average child’ at a given age. These tasks included tests of linguistic ability, such as naming objects and following commands, tests of memory such as remembering a string of digits, and tasks based on reasoning and deduction, such as pattern-completion exercises. Taking the view that a child’s mental capabilities increase with age, Binet developed a scale for analysing the results of his test, based on the hardest problem a child could solve: for instance, a child of, age 7, who could successfully tackle tasks aimed at the ‘average 7-year-old’ could be assumed to be of average intelligence for their age. However, a child of 7 who was not able to address tasks harder than those aimed at five year olds for example, would be identified as less mentally able, and hence potentially in need of additional educational support, and vice versa. The test result was expressed as the child’s ‘mental age’, the highest age level at which a child was able to complete most of the tests. Published in 1905, the test became known as the Binet-Simon Scale. It rapidly gained popularity, and became widely used in both Europe and the US.

As the scale became more widely used, so it began to undergo revisions. Lewis M. Terman, a psychologist at Stanford University, had an important part to play in this. Binet’s original intention may have been to identify underperforming school pupils, but Terman expanded the test’s scope: he extended the test’s age range into adulthood, and began to use it to identify children and adults who were unusually intelligent relative to their peers, as well as those who were struggling. Terman also did away with ‘mental age’ as a measure of intelligence, instead developing a more intuitive way of presenting the results of the test (by now known as the Stanford-Binet Scale): Dividing mental age by chronological age and multiplying by 100 yielded a scale in which a child of average intelligence for their age would score exactly 100. Terman named this score the ‘intelligence quotient’ or ‘IQ’, as it is ubiquitously known today. Though very effective when applied to children, this method presents a difficulty when attempting to calculate a score for adults: for most of adult life, one’s ‘mental age’ would not be expected to change: put simply, a 37 year old would be expected to have much the same in terms of mental capabilities as a 35 year old, or a 39 year old.  This problem was overcome by Wechsler, an American psychologist who worked extensively in the field of intelligence testing. He suggested that the IQ ratio, as previously calculated, should be replaced by a ‘deviation score’. In other words, test subjects were assigned a score based on how differently they performed relative to other participants of their age, based on the statistical spread of all test scores for a given age group. This method caught on, and is used to this day.  Wechsler also examined the various subtests that were in use, modernising some of the older ones and replacing them with more up-to-date versions. The final test included subtests for both verbal and performance (non-verbal) skills. Gradually, the Wechsler Adult Intelligence Scale (WAIS) took form. Modern IQ tests are almost exclusively based on revised editions of the WAIS, which is now in its fourth edition (WAIS-IV). There is also a counterpart for test subjects between the ages of 6 to 16, the Wechsler Intelligence Scale for Children (WISC).


Throughout their history, IQ tests have purported to measure the same quality. However, as IQ tests have evolved, so has our concept of what constitutes intelligence. In their early days, despite the tests consisting of several different subtests, the focus was on generating a single overall intelligence score. This was a product of the prevailing theories of intelligence at the time: Charles Spearman, an English psychologist and statistician, noticed that children who performed well in one school subject tended to perform well in all other subjects as well. This led to Spearman suggesting that that a single global entity underlay mental performance in all areas - Spearman termed this ‘general intelligence’, or ‘g’. I.Q tests appeared to bear out this theory: a test subject’s score for each subtest tended to correlate with their score for each of the other subtests, as well as their overall score for the intelligence test: someone who performed well in one subtest was likely to perform well in all of them, and hence perform well on the IQ test overall. Furthermore, a test subject’s results for different IQ tests also tended to correlate.

Nevertheless, the concept of ‘g’ was unable to account for the fact that the correlations between certain subtests were stronger than others. In fact, subtests appeared to form ‘clusters’ of correlations - perhaps intelligence was not a single entity, but in fact comprised of several abilities? Most notably, ‘verbal’ and ‘performance’ tests appeared to form separate clusters. Interestingly, it has been found that certain disorders characterised by low IQ seem to affect the scores in one cluster more than the other. This leads to the possibility that the various components of intelligence are distinct and arise separately. However, the concept of ‘g’ cannot be entirely discarded - invariably, all ‘clusters’ are affected in such disorders; the deficit is simply more pronounced in some clusters than others. In practice, modern theories of intelligence are shaped around a three-level hierarchy: at the bottom, numerous basic factors underpin performance on individual tasks. Higher-level, intermediate factors account for performance in task clusters. Finally, at the apex of the hierarchy, a single factor, ‘g’, determines overall mental ability and performance. This theory is taken in to account by the WAIS-IV: as well as generating a single overall score, subtests are grouped into four categories (Verbal Comprehension Index, Perceptual Reasoning Index, Working Memory Index, Processing Speed Index), for each of which an individual score is also generated.


Wechsler defined intelligence as ‘the aggregate or global capacity of the individual to act purposefully, to think rationally and to deal effectively with his environment’. But just how good are IQ tests at measuring such a capacity? Fundamentally, despite the multiple subtests which supposedly cover a wide range of abilities, the modern I.Q test takes a rather narrow, academic view of intelligence, focusing on paper-based problem-solving. Therefore, it becomes necessary to find out whether IQ has any bearing on some other, independent measure of ‘effective interaction’ with one’s environment. In research, such measures have included level of educational attainment, and occupational success. These measures correlate with IQ, but also with a wide variety of different factors, such as, parental socio-economic status, neighbourhood, birth order, and number of siblings. Nevertheless, even once these factors are accounted for, IQ is still a good predictor of overall success. It is worth noting that this relationship is, however, not absolute: Wechsler himself observed that one of his tests subjects who scored a mental age of 8 on the Stanford-Binet scale ‘had gotten along very well, was supporting a family, had been working as a skilled oil-driller for several years and was earning from $60 - to $75 per week’. This would suggest that IQ does not fully account for all the aspects of the definition of intelligence.

So what other factors might be able to account for ‘good interaction with the environment’ in an individual of low, or even average IQ?  And are there any areas in which an individual of high IQ might stumble? One place to look might be the culturally pervasive stereotype of the highly intelligent professor, who is nevertheless absent-minded and socially inept. Many advocate the existence of a ‘social intelligence’, which in contrast to ‘general intelligence’, as measured by IQ, governs a person’s behaviour and decision-making in a social context. This quality is often likened to ‘common sense’. Though popular culture would have us believe that social intelligence and general intelligence are distinct, or even mutually exclusive, this view is not corroborated by research: performance on an abstract reasoning task similar to an IQ test was predictive of performance in a social reasoning task. However, the relationship was not entirely straightforward: of those who solved the abstract reasoning task, all solved the social reasoning task - but of those who did not solve the abstract task, some were still able to solve the social reasoning task. Therefore it is not implausible that these two hypothesised types of intelligence should be, at least in part, distinct. More broadly, the focus on social intelligence comes in the wake of a movement supporting the existence of ‘multiple intelligences’. It is suggested that these cannot all be measured in the context of a pen-and-paper IQ test, which draws very heavily upon the use of logic and linguistic skills. Furthermore, there is a swing away from the view that a high IQ in isolation is a desirable attribute, particularly in work environments which require high levels of communication and cooperation - rather, other proposed intelligences are seen as potentially more fundamental due to their proposed potent effect on interpersonal interactions.


So is a high IQ sufficient for a successful presidency? Estimates of the IQ scores of U.S. presidents generally place them on the high side of the spectrum, and the perceived effectiveness of a leader certainly appears to depend on their IQ. However, IQ appears to be only a small piece of the puzzle: for instance, ‘emotional intelligence’ appears to have an important role to play in the effectiveness of domestic leadership: somewhat akin to ‘social intelligence’, emotional intelligence is a measure of one’s ability to recognise and respond appropriately to emotion, using this awareness to guide one’s thinking and decisions. Others have suggested that even more nebulous measures, such as ‘cultural intelligence’ (the ability to understand and be sensitive to cultures other than one’s own, behaving appropriately in situations characterised by cultural diversity), are particularly important for modern heads of state, particularly in the context of international leadership. Interestingly, a high IQ is associated with personality traits such as ‘Openness to Experience’, which predispose to more liberal views, and hence a greater tolerance for personal, political and cultural differences. Therefore it could be suggested that IQ is a contributing factor to ‘cultural intelligence’, bringing us back to the idea that IQ is a key underlying factor of leadership qualities.

As with all subjective qualities, leadership effectiveness is hard to measure objectively, with research often relying on questionnaire answers from focus groups. As such, it is difficult to calculate the extent to which any given factor affects this quality. However, a general consensus appears to be developing, which falls in line with what many would view to be a common-sense conclusion: an effective leader is not simply someone who is smart: they are someone who has a good understanding of people and their interactions.

So does Donald Trump have a ‘high IQ’, as he relishes to claim? Perhaps. Does he have the qualities required of the leader of one of the world’s biggest international players? A simple IQ test almost certainly does not hold the answer.


Add a comment