Recent studies investigating the relative roles of nature and nurture in education have provoked much debate. This debate is partly fuelled by a misunderstanding, or miscommunication, of the term ‘heritability’ and its scientific meaning. Getting to grips with key scientific terms will help to move the debate forward and encourage productive engagement between education and science.

Research is making great strides in uncovering the role of genes in educational outcomes. With the arrival of more research it is essential that non-scientists, including educators, parents, and policymakers, understand what the terminology means so that they can meaningfully engage in debate.

At present, considered debate about how to improve educational outcomes, particularly for low performers, can be hindered if we don’t accept scientific findings that appear to contradict our values or intuition. In part this is due to misunderstandings of scientific terms, and one term that has a particular tendency to be counterintuitive, and has caused much confusion and debate, is heritability.

For a given trait or educational outcome, heritability describes the proportion of variation in the population that is accounted for by genes. To take an example, GCSE (General Certificate of Secondary Education; the tests taken at age 16 in the UK) results were estimated to be 62% heritable. This is sometimes incorrectly taken to mean that 62% of an individual’s GCSE result is explained by their genes. What it actually means is that 62% of the differences in GCSE results between members of the population is explained by genes.

Heritability estimates depend on the environment. If the environment is very similar for everyone, then heritability of a trait is likely to be higher, since individual differences in the trait can’t be explained by different environments. In contrast, if environments are very different for different members of a population, then the heritability estimate is likely to be lower, since different environments will explain more of the variation.

As a result, heritability estimates can differ between countries due to different environments. For example, a country with a very standardised school system might have much higher estimates of heritability compared to a country where schools vary widely.

“In addition to changing depending on the environment, heritability estimates also change with age.”

A very high heritability estimate, therefore, does not mean that the environment does not matter, it just means that in this specific population, genes explain more of the variation. For GCSE grades in the UK then, the fairly high estimate of 62% heritability should not be taken to mean that schools provide little value; rather it may suggest that UK schools are providing similar learning environments.

In addition, a heritability estimate does not tell us which specific genes are associated with the differences we see. It therefore cannot tell us how much of an impact an individual gene has, or the mechanism through which the genes are having an effect. Studies are now starting to investigate specific genes, and the evidence is showing that many genes have an impact on traits and educational outcomes, with each one of these genes having a very small impact.

In addition to changing depending on the environment, heritability estimates also change with age. Heritability estimates tend to increase with age, such that more variance is explained by genes than environment as we get older. One estimate is that genes explain 41% of the variability in intelligence in children, but 66% of the variation in intelligence in young adults.

“Schools and teachers have a crucial impact on students’ learning, and high heritability likely reflects the fact that all students sampled are exposed to high quality learning environments.”

This is a counterintuitive idea since we might expect that as we have more experiences we get further from any genetic influences. But to the contrary, heritability is thought to increase because we have more choice over our experiences as we get older, choosing to engage in activities that suit our genetic propensities.

Heritability is a complicated term, with a very specific meaning that can be misinterpreted. Perhaps the most important thing to take away is that a high heritability estimate does not mean that the environment is not important. Schools and teachers have a crucial impact on students’ learning, and high heritability likely reflects the fact that all students sampled are exposed to high quality learning environments.

“With a shared understanding of what heritability means, we can start to have meaningful debates about what we want our education systems to achieve.”

As our understanding of the role of genes in learning increases, it is essential that discussions are based on common terminology. Heritability is a good place to start. With a shared understanding of what heritability means, we can start to have meaningful debates about what we want our education systems to achieve.

Heritability estimates tell us:

  • The variation in a certain trait or outcome measure that is explained by genes
  • About the genetic influence in a given population

Heritability estimates cannot tell us:

  • Which genes are important for a trait or outcome measure
  • How genes have an impact on traits or outcomes
  • The role of an individual’s genes


  1. I read your piece for npj Science of Learning entitled ‘Do Our Genes Determine Learning Ability’ and this led me to your blog. I am already benefiting enormously from having done so. My interest in your work stems from over thirty years in teaching, mainly in primary schools, and a consuming interest in how to make learning more effective for young people of all abilities. I was taken by your remark, “Scientists should engage with those who are affected by their work”. I wholeheartedly agree that they should, but am finding it virtually impossible to generate interest among researchers working in the education field in an area which seems blighted by any mention of cognitive ability, namely attainment.

    I am currently engaged in a small research project in collaboration with a retired secondary school headteacher, Roger Titcombe, author of ‘Learning Matters’. We are looking into the vexed subject of social mobility, specifically in relation to the alleged north/south school attainment gap. I agree with you that “The best way to proceed is to open the conversation and consider the most desirable way of incorporating research findings into practice.” However, you rightly question, “why is genetics such a controversial topic in education?”, to which there is no logical answer. Whether in dealing with teachers, school managers or a whole host of education professionals this is a no-go-area. The fear of inviting eugenics into any debate, despite a resolute insistence that determinism never has and never can have any part to play, seems to paralyse thought and drive people in the opposite direction.

    You write, “Teachers and schools are judged on the performance of their pupils, and as such aim for the highest grades for all students across subjects, yet a genetically informed approach might lead to greater acceptance of differences between pupils.” Your final conclusion is important for it is the only way we have of eventually convincing the wider world that being different is a challenge (in any sphere) only if we misunderstand how to approach it. The school system, as it is currently working, runs directly counter to this, to such a degree that your comment about schools aiming for the highest grades for all students across subjects is inaccurate.

    Schools are indeed judged on the performance of their pupils but this is the result of totally misguided political interference in narrowing down what is taught and restricting learning to coaching and the memorisation of facts. This is depriving young people of the kind of education that would better suit their needs. This is not education. By refusing to accept the heritability of traits, too much emphasis is put on the prescriptive tweaking of certain environmental factors selected to bring about the impossible, namely aiming to increase the number of students gaining average or above attainment in defiance of the natural curve of distribution. The outcome is the exact opposite of what you advocate, that “teachers should continue to provide the best educational practices,” whatever these may prove to be.

    1. Thank you so much for your interesting and thoughtful comments John, it’s really great to hear that you have benefited from reading my posts. I’m pleased that you agree that researchers should engage with teachers, and I’m disappointed to hear that you haven’t had any luck with researchers on the topic of attainment.

      I’m involved in a Wellcome Trust funded online project connecting researchers and teachers called the Science of Learning Zone. I’d recommend having a look on the website ( where you will be able to find researchers who are keen to engage with teachers. The site allows you to ask questions of researchers (anything you like), so perhaps there is something you could ask that might generate interest among enthusiastic scientists.

      Your research project on social mobility sounds fascinating and I look forward to seeing the results. With regards to genetics, as mentioned in the post above, I really think a large part of the problem is individuals not understanding what results reported in the media mean (through no fault of the individual). Often the intuitive interpretation is wrong, which means that real discussion can never happen because the results of genetics studies are rejected.

      Thank you for raising the final point about the current systems not allowing teachers to provide the best possible education. It of course reminds me of Michael Gove’s comments about expecting all schools to be above average.

      Thanks again for getting in touch John. I have never been a teacher so am always interested to hear from those who can speak more accurately about the realities of being in a classroom every day. Wishing you all the best in your future teaching.

    2. This is a very interesting point that you raised John. I am currently starting to study genetics and development in graduate school and had recently been surprised by the kind of ire that bringing up research on heritability of intelligence seems to inspire. From the unhappy history of eugenics it appears to have become a forbidden zone – the reason give is that policy-makers and teachers cannot be trusted with scientific facts. Policy-makers because apparently they will choose to not devote any resources to groups of children, and teachers because they will feel that it does not make sense to try to teach certain students. This shows a great deal of mistrust of others in a society. While I do not think is entirely unfounded given the negative news that comes out… but personally, I believe that closing the conversation among those who mean well does not prevent such information from being used by those who do not mean well.

      I find such an article like this one really wonderful for explaining the nuances of the results. How heritability is not determinism. And that in fact, such information only points more to our environmental affordances as they fit children – how individuated they can be. I am curious what this implies: Could an environment with more heterogeneous, open opportunities give children the chance to develop to the best that they could be (reducing heritability effects)? Please correct me if I am wrong in this interpretation. I would love to have this elaborated upon.

      From personal experience, I grew up in Taiwan, where I consider the educational environment very homogenous. Our education system is very academic (testing) focused, where tracking into vocational school is the only option for those who cannot get good enough grades. In addition, there is a myth as to the ‘meritocracy’ of our system : if you test well, it is because you worked hard. If you don’t test well, you have been lazy. Recently, one of those kids who had gotten the perfect score for college entrance exam confessed in an article that he had told the reporters he studies hard and smart… etc. When actually, he had basically a photographic memory, so never really needed to study. He only said those things because that was what the reporter wanted to hear. In the converse, I have several friends who went to high achieving schools and felt utterly crushed by constantly trying to compete grade-wise with their wunderkind classmates. Only later on (generally after they started working and were independent) did they begin to feel free to explore their interests which had been previously deferred.

      I see the potential in this sort of research in giving children what they need, and diversifying the school offerings (and evaluations in meaningful ways). If the assumption is good grades will result if a teacher is skilled enough, and the child works hard enough, then the academic grading makes sense. However, if the assumption is that there are certain children who will test exceptionally well because they are naturally gifted, then using academic performance only to evaluate teacher and student performance really doesn’t make as much sense. Because our current system operates under the myth of this equal outcomes with effort, the bar is set at where the most high-performing children are at to differentiate for evaluation and tracking. This is unreasonable, unrealistic, and results in a soul-crushing competition for some children while elevating others for only ONE aspect of their talents. I believe this results in class differentiation. And inadequate resources are put into fostering and valuing other existing talents and interests, both for those who are not gifted with high IQ, and for those who are. There are many people in Taiwan who have tested extremely well, and never developed other important aspects of their personality (such as citizenship or awareness of how to handle relationships).

      I believe that ignoring the fruits of intelligence research results in a similar blindspot for our public conversation about policy design and citizenship the same way ignoring teaching children to be color-blind by refusing to acknowledge race causes a problem – where there is blindness, assumptions (often wrong ones) come up to fill the void. And the outcome can be worse than what we had imagined.

      Lastly, I would like to mention something interesting I’ve come across: in the Netherlands, there is a diagnosis for borderline intellectual disability. This is not a distinction that is common for every country. The Dutch government uses this diagnosis to offer support for people who have this issue, because even though individuals who have (under this criteria) only borderline lower IQ, they do suffer many functional difficulties – in daily living, finding and maintaining jobs, partnership, parenting…etc. For example, a representative can go to an employer and explain the difficulties that the individual suffers in their life, their strengths, and work with the employer in devising a strategy where the individual can work productively so they can stay employed. I thought it is a very pragmatic way of looking at this: it is after all in society’s interest that people are not unemployed. In this case, such a diagnosis is not a stigma, it’s an opportunity to receive the help that is needed in a way that is useful for the individual and a worthy investment for society.

Comments are closed.

Keep up to date with the BOLD newsletter