Issue 3: School achievement in literacy and numeracy: Insights from research using the twin design

ISSN: 2652-5518
By Brian Byrne

Main points 

  • Twins studies show that genes explain half of more of the variability in children’s achievements in literacy, language and numeracy …
  • … but genes are not destiny! Environmental influences can be very important, especially for disadvantaged groups.
  • To lift achievement levels most effectively, funding models should be student-directed rather than school or class-directed.


The aim of this article is to summarise what twin research has taught us about Australians’ school achievement, and to outline the implications for students themselves and their families, for teachers and educational policymakers, and for the broader public as it debates education and its funding. Here I focus on the foundation skills of literacy and numeracy.

Suggested citation: Byrne, B. School achievement in literacy and numeracy: Insights from research using the twin design. Conversations in Twins Research, Twins Research Australia, Melbourne, 2018,

Nature or nurture: where do differences in school achievement spring from?

In an ideal educational system, all children would read and write to a high standard. The same would hold for the range of numeracy skills. Then the trajectories that individuals follow throughout their schooling and beyond would not be held hostage to poor literacy and numeracy; students would be able to develop and indulge individual interests in social sciences, languages, physical and biological sciences, art, music, etc. It would even allow changes in direction from, say, the humanities to the sciences if someone has a change of heart later in their academic progress.

Clearly, experience tells us that it is an unrealistic goal to expect the foundations to be equally and highly developed in all school students. Despite sympathetic and often individualised instruction, some struggle with learning to read and write and with mastering numerical skills. These struggles can project through schooling, undermining aspirations and narrowing academic and life opportunities. We would like to know why this is true, and we would like to know what can be done to help these students reach acceptable levels of competence. Equally, we would like to know why some students seem to breeze through the basics of literacy and numeracy and whether we can harness these processes for the benefit of all.

This is where data from twins can make a substantial contribution. My colleagues and I have harvested results from twin students and their siblings who have sat the tests referred to as NAPLAN (National Assessment Program: Literacy and Numeracy, administered in Years 3, 5, 7 and 9) [1]. So far we have had the cooperation of over 3000 families Australia-wide, and used the data to estimate the relative roles of genes and aspects of the environment in influencing test scores. We had earlier conducted a twin study of literacy starting with preschoolers, following their progress until Year 2, a four-year span in total [2].

What the research tells us

Here are some lessons from our twin-based research.

Genes matter for relative attainments in literacy, language and numeracy. The table below shows the percentages of variability in NAPLAN test results averaged over the four year levels. A represents genetic influence, C represents environmental factors that affect both twins in a family, and E represents environmental factors that affect one member of a twin pair separately from the other.

 Percentage of variability

Genes (A) account for half or more of the variability among Australian school students in the four language skills and in numeracy across these middle years of schooling. The environment that twins share (C), such as the home, the school and, sometimes, the teacher, is the least influential of all sources. The environment that twins do not share (E), such as separate friends, illnesses, the intrauterine environment and, sometimes, separate teachers, account for about a fifth to a half of the variability.

A similar pattern holds for literacy even when assessed as early as the end of kindergarten. Using word-reading data collected on kindergarten twins in the Sydney metropolitan area during our earlier study, we estimated the proportions of variability in results explained by A, C, and E at 79%, 6% and 15% respectively [2]. The influence of genes on reading is seen early in the school experience, because it can be reliably measured.

This pattern holds in other countries and for other subjects. Studies from Australia and around the world have confirmed genetic influence of between 50% and 80% for literacy and numeracy, and for other school subjects such as science and second-language learning.

Research with twins can also help us understand how genes and the environment influence a trait. For example, in our research with twins in kindergarten to Grade 2, we have shown that the process of storing mental images of newly-encountered written words varies from child to child and that this variability is driven in part by genetic differences. I draw out some conclusions of this finding later, in the section “Implications for teachers”.

Some caveats

Here are some things that need to be borne in mind concerning our results.
Environmental influence matters even though genes matter. This is true in at least two ways. First, for each NAPLAN measure genes are only part of the story; you can see in the table above that 25% (spelling) to 54% (writing) of the differences are under the influence of environmental factors. The child’s experiences clearly pay a role, and because it is those that parents and teachers can help shape, it is important that we understand them as fully as possible. Second, for some groups within society the environment can be particularly important. For instance, there is evidence that children living in areas with heavy metal contamination of the air and soil can have depressed NAPLAN scores, and the lower-than-average school attendance evident among some Indigenous children can lead to lower school performance.

The relative roles of genes and the environment derived from large numbers of twins cannot be projected onto the individual child. It makes no sense to say, for instance, that genes contribute 60% to a particular individual’s NAPLAN reading score, factors in the environment shared within a family 11%, and factors in the environment peculiar to an individual 29%.

Genes are not destiny. A substantial role for genes does not mean that a child’s level of achievement cannot be lifted. We know from medical interventions that even diseases that are almost completely determined by genes can be ameliorated. The metabolic disorder phenylketonuria is a prime example; early identification followed by dietary restrictions and supplementation normally allows the sufferer to live a normal life. Marked reading disability (dyslexia, as it is sometimes called) is another; there is ample evidence that well-designed and well-delivered interventions, combined with family support, can help children reach normal levels of reading.

Implications for families

Avoid self-blame. If, despite normal school opportunities and a supportive home environment, a child is struggling with literacy or numeracy, it is possible that genetic factors are at work. Understanding that can be something of a relief in that difficulties with reading are not likely to be due, for example, to parents failing to read to this child in some “proper” way.

Avoid pessimism. At the same time, families might be tempted to feel that there is little that can be done the help their child, given that genes may be playing a role. But, as noted earlier, genes are not destiny. This message is important so as to avoid the temptation to give up. Professional help is available, and is known to work.

Implications for teachers

Recognise that genes matter. Simple recognition of a substantial role for genes in academic progress can and should shape expectations about how easy it will be to help children struggling with literacy and numeracy. It will not just be a matter of tinkering with a child’s circumstances – with, say, placing them in a smaller class, or providing them with a tablet computer. While, as stated earlier, we cannot determine the degree to which a particular child’s progress is affected by their genetic endowment, we can say that in the school system as a whole we cannot expect environmental engineering (class size, digital access, etc.) to go very far in solving the problem of struggling learners. It can help, of course, but is unlikely to be the panacea that some enthusiasts contend.

Take advantage of progress in understanding how genes and the environment affect school achievement. Earlier I mentioned that processes governing how written words are stored in memory are partly under genetic control. This means that children who carry lesser numbers of the relevant (though unknown) gene variants are likely to need to see words more frequently than other children in order to add them to their stock of automatically-accessible written words. In turn, this means that the very thing that struggling readers are likely to want to avoid, the act of reading, needs to be encouraged in them so that they accrue the extra practice. Figuring out how to do this is part of the high art of teaching, and many sensible ideas have been proposed (ensuring that reading material appeals to the children, providing books and magazines for children whose families don’t have them, and so forth).

Avoid self-blame. Any teacher worth their salt will be alert to their own strengths and weaknesses, and be on a path to increase strengths and abolish weaknesses. But having adopted this stance, teachers, like families, can take some comfort in knowing that a particular child’s problems can have their origins in genetic factors, something that they, the teacher, has no control over whatsoever.

Avoid pessimism. Genes are not destiny. Keep that in mind.

Implications for educational policymakers

In designing employment and remuneration practices for teachers, bear in mind the role of genes in students’ achievement. It is commonplace for journalists and politicians to point to teacher characteristics like dedication and qualifications in explaining poor school results. However, our results say that genes play a substantial role, though to an unknown degree in any particular classroom or school situation. Therefore, denying advancement or even continued employment to teachers whose classes fall behind is, in our view, unjust. Of course, poor teaching amounting to malpractice needs to be detected and corrected, just as it does in medicine. But just as we do not routinely blame doctors for, say, a local cluster of cancers or outbreak of measles, so we should not routinely blame teachers for a decline in school performance in a region or in a class within a school.

Funding models designed to compensate for low levels of performance should be student-directed rather than school- or class-directed. This advice flows from the pattern of results in the table presented earlier. Eighty-five per cent or more of the variability in each NAPLAN domain is down to a combination of genes, A, something unique to individuals, and environmental factors unique to the individual, E. The rest, environmental factors like the school, class or family – things that twins share – account for a maximum of 15%. Thus, funding models that “stop” at the school level can be siphoned off into group-level things like smaller classes or more computers, and barely reach the individual who needs the extra support that the funds could underwrite. Ways need to be found to direct the funds to individual struggling students, including those who happen to be in schools that as a whole perform well.

Our results have implications for funding levels as well as for funding delivery. In the hands of some commentators, evidence for a role for genes in school achievement is turned into an argument that it is pointless to increase funding for schools because it won’t help – genes are destiny, goes the argument. But as we’ve said, genes are not destiny, and this kind of argument needs to be nipped in the bud. We do not say that because genes are implicated in many diseases we should constrain funding for the health system on the grounds that there’s nothing we can do about those particular diseases. We argue the opposite, that health funding should be as high as economically sustainable because treating those diseases may indeed require more strenuous efforts than ones for which the environment is a big factor and therefore open to easier modification. The same is true for education funding.


I hope that readers can see the value of twin studies in the important area of school achievement. Twin studies not only allow us to identify factors that influence school performance, but provide a basis for recommendations about educational practice and policy.

About the author

Brian Byrne is an Emeritus Professor in the School Psychology, University of New England, and a member of the ARC Centre of Excellence in Cognition and its Disorders as well as of the NHMRC Centre of Research Excellence in Twin Research. His PhD is from McMaster University, Canada. His research interests have centred on literacy and language development, most recently using data from twins who have sat the Australia-wide NAPLAN tests, as described in this contribution.

If you have any comments or questions, please email us at


1. Grasby, K.L., W.L. Coventry, B. Byrne, R.K. Olson, and S.E. Medland. Genetic and environmental influences on literacy and numeracy performance in Australian school children in Grades 3, 5, 7, and 9. Behavior Genetics, 2016. 46:627-648.
2. Byrne, B., S. Samuelsson, S. Wadsworth, J. Hulslander, R. Corley, J. DeFries, P. Quain, E. Willcutt, and R. Olson. Longitudinal twin study of early literacy development: preschool through Grade 1. Reading & Writing, 2007. 20:77–102.

Further reading

Byrne, B., S. Wadsworth, K. Boehme, A.C. Talk, W.L. Coventry, R.K. Olson, S. Samuelsson, and R. Corley. Multivariate genetic analysis of learning and early reading development. Scientific Studies of Reading, 2013. 17:224-242.

Haslam, N., and E.P. Kvaale. Biogenetic explanations of mental disorder: the mixed-blessings model. Current Directions in Psychological Science, 2015. 24:399-404.

Olson, R.K., J.M. Keenan, B. Byrne, and S. Samuelsson. Why do children differ in their reading and related skills? Scientific Studies of Reading, 2014. 18:38-54.