How often do you consider your students’ intelligence quotients (IQs)?

It’s not a measure we refer to a lot in schools, but it’s one we have all heard of - and it’s at the root of how many of us think about intelligence, or academic “ability” more generally.

Within the realm of education, IQ is often presented as a measure of biological potential; just as inches measure height and kilograms measure weight, IQ measures intelligence. 

But is this accurate? And if not, what does that mean for how we think about students’ potential?

To get to the bottom of this, it’s important to first understand exactly what IQ is and how it’s measured. 

IQ tests are composed of a hodgepodge of random, unrelated tasks. For instance, a test might pose the following questions:

1. Complete the following numerical pattern:
   5, 10, 15, 20, ???
   (Answer: 25)
    
2. Can the following letters be rearranged to form an English word?
   T O O M T
   (Answer: Yes - “motto”)
    
3. What four-letter word can be added to the end of these to make brand-new words?
   Cheese  -  Cup  -  Fruit  
   (Answer: “cake”)

An individual’s answers to these varied questions are amalgamated and then used to assign an IQ score - typically between 40 and 160, with 100 being average. In fact, the average IQ score in the UK just so happens to fall smack dab on 100. 

Using this as a reference, what do you think the average IQ score is in New Zealand? Or in China? Or in Nigeria? 

The answer, for all three countries, is 100 - because the average IQ in every country around the world is 100. How can this be? 

The reason is that IQ does not reflect a person’s actual performance on the test. Rather, it represents a rank of where you stand, compared to similar people who took the test around the same time as you.

To explain this in a bit more detail, imagine if everyone in the UK took an IQ test and failed miserably, getting no answers correct. Now, imagine you took that same IQ test and got one correct answer. Seeing as the average performance was zero, everyone in the UK would earn an IQ of 100, while you (since you scored above average) would earn an IQ of around 130.

The same would hold true if this situation happened in reverse. If everyone else in the UK took an IQ test and got 100 per cent of the answers correct, and you took that same test and got just one answer wrong, you would earn an IQ score of around 70 (because you performed below average), while everyone else would earn an IQ score of 100.

In the first example, you get one correct answer and look like a genius, while in the second example, you get many more correct answers but look like a fool.

This leads us to two very important considerations.

First, IQ scores can never be used to determine an individual’s actual ability. Consider if we applied the same approach to height: I could rank a group of people from shortest to tallest, but knowing that you’re the tenth tallest tells me nothing about how tall you actually are.

Second, some people argue that IQ doesn’t change across a person’s lifespan. If true, this simply means that a person’s rank doesn’t change, not that intelligence remains the same. Let’s return to height: if you’re the tenth tallest when you’re 5 years old, there’s a good chance you’ll be the tenth tallest at 45 years old, too. But this doesn’t mean you won’t grow any taller across those 40 years.

The Flynn effect

Is it true that even a person’s IQ rank is destined to remain stable throughout their lifetime, though? 

In 1990, researchers put this question to the test. They asked adults to undertake four minutes of explicit training on a short-form IQ test called Raven’s Matrices. They were shown a single strategy designed to help them perform well in the test, and allowed to practise it using five examples. 

One week later, these adults showed a 13 per cent increase in performance on the IQ test. By comparison, a group that did not receive training only showed a 3 per cent increase. Just four minutes of practice, then, was enough to significantly change IQ ranking.

This is far from an isolated case. More than seven decades of research have demonstrated that when people explicitly learn about IQ tests and practise various strategies, their rank increases - something that would be impossible if IQ tests were measuring a biological trait immune to intervention.

But if IQ tests aren’t measuring an innate ability, what exactly are they measuring?

The work of intelligence researcher James Flynn helps to explain. He spent years exploring and documenting what has become known as the “Flynn effect”: a phenomenon in which steady improvement in performance on intelligence tests has been observed in many countries over a number of years.

To illustrate this, imagine if we took a person of average IQ according to the standards of 1922 (IQ testing was invented in 1904), and re-ranked them according to the standards of 2022. That person would now have an IQ score of 70, rather than 100.

If we took a person of average IQ by today’s standards, on the other hand, and re-ranked them according to the standards of 1922, then that person’s IQ score would jump from 100 to 130.

Remember: IQ doesn’t measure actual performance - it compares and ranks people who took the test around the same time. This means the average IQ for every generation is 100 - and yet, in terms of actual performance, every generation has gained about six IQ points on their parents.

So, what is happening here? The explanation is simple: school.

‘School ability’

In every country studied, when mandatory schooling comes into effect and the number of years each child spends in formal education increases, so too does IQ. In fact, for each year of school children attend, they gain between one and five IQ points. 

This is because IQ tests aren’t measuring some form of general intelligence - they’re measuring a very specific form of intelligence that is reflected in traditional education. Furthermore, just like any skill, this “school ability” improves with practice.

Evidence for this comes from a very unexpected place.

You’ve no doubt heard the stories: identical twins, separated at birth, reunited 50 years later to find they like the same food, enjoy the same music, and both have a dog named Spot.

If IQ was measuring innate intelligence, then separated twins should share the same IQ rank regardless of circumstance. So, is this what we find?

Believe it or not, despite all the hype surrounding this type of research, individual IQ data has only been published for 91 pairs of separated twins to date (aggregate IQ data has been published for several hundred additional separated twins - unfortunately, aggregate data can not be analysed for moderators). Of those 91 pairs with individual IQ data, only 34 experienced significantly different forms of education - 57 grew up in the same town and attended schools within the same district. 

Those 57 twin pairs who experienced near-identical schooling differed on average by only six IQ points. For context, this is the same average difference that we see in pairs of identical twins raised together. 

However, those 34 twin pairs who experienced significantly different schooling differed, on average, by a whopping 13 IQ points - meaning they were closer to complete strangers (who differ on average by 16 points) than identical twins.

As schooling changes, so too does IQ. 

But if what IQ measures is really some kind of “school ability”, rather than a general form of intelligence, then what exactly is intelligence?

What is intelligence?

Crows are incredibly intelligent. In fact, there is research showing that they will add stones to a pitcher of liquid in order to raise the water level to gain access to floating food. According to psychologist Jean Piaget, this is something most human children won’t consider until they are between 5 and 9 years old.

Why does this matter?

Many people believe that IQ tests measure biological intelligence. Unfortunately, when we give crows an IQ test, they score a whopping 0.

This is because intelligence isn’t something physical we can measure; it’s not a biological entity that can be located and isolated, like the heart or the lungs. Intelligence is not something a person has, it is something a person does. In other words, people aren’t generally intelligent - they act intelligently in specific contexts.

Once we recognise intelligence as a contextual behaviour, we are free to acknowledge all its myriad manifestations. Everyone - even a crow - has the ability to learn and act intelligently inside particular contexts without the need for some general, all-purpose biological trait that manifests equally across every circumstance.

Consider the fact that in the UK at the turn of the century, hundreds of children were being raised on canal boats, with no formal schooling and little social contact beyond other canal-boat families.

When these canal-boat children were IQ tested at a young age, they were close to average (scoring 90 at age 5, on average). It was only as children grew older that their rank started to drop significantly (to a score of 60, on average, at age 18).

If intelligence is biological, then these children must have lost intelligence the same way people lose weight.

However, if intelligence is any meaningful behaviour within a specific context, then we can freely recognise their incredible navigation, engineering, fishing, weather forecasting and memorisation skills and call them what they rightly are: intelligent.

Human beings grow and adapt to the environment they are surrounded by. If a group of people are not surrounded by traditional schooling, then there is no reason to assume they would ever develop those specific “school ability” skills measured by IQ tests.

In the end, human intellect isn’t powerful because it’s some specific, all-purpose trait. Human intellect is powerful because it’s highly adaptive - any healthy human being can demonstrate intelligent action within whatever realm they dedicate their efforts toward.

So, what does this mean for education?

First, using IQ - or other tests designed to measure general intelligence - as a stratification tool may be questionable. Once we remember that intelligence is an action rather than a trait, it makes far more sense to use actual student behaviour as a tool to drive stratification.  

Second, it’s important for teachers and students to recognise that IQ is neither a true measure of intelligence nor a fixed entity. As with every skill, the ability to excel in school can be improved with practice. In other words, school ability (IQ) will grow or shrink depending upon the amount of time spent and effort expended within a school setting.

Finally, for those who fear an average or below-average IQ score is a guarantee of mediocrity, it is not.

As psychologist Dean Keith Simonton - a man who has dedicated over 40 years of his life to researching exceptional human performance - clearly states: “A high IQ is simply not required for genius-level behaviour, and even a low IQ is an exploitable resource you can use to achieve genius-level behaviours within a particular field.”

All of this is perhaps worth bearing in mind as we enter the summer exams season. While Sats, GCSEs and A levels aren’t directly based on IQ, they do strongly correlate - meaning these exams measure a similar “school ability” definition of intelligence.  

This isn’t a bad thing: school ability is an incredibly important skill and it might be worth knowing where our students rank on this measure. However, it’s important we recognise that this doesn’t constitute the full nature of intelligence and that these tests can’t provide a complete picture of a student’s ability or capacity.

Jared Cooney Horvath is a neuroscientist, educator and author