After all the PISA hype, here is a study that is only indirectly related to education! A massive new study in Nature provides important new insights into the link between our personality and our genes. For this, the (truly ridiculous number of) researchers combined data from 46 cohorts. The number of participants ranged from well over 600,000 to more than 1.1 million, depending on the personality trait (no, that is not a typo). They examined the Big Five. This is probably the most important current theory on personality: extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience. Furthermore, these can influence how someone behaves in the classroom. The OECD has even used the Big Five as a starting point for its work on social and emotional skills. Therefore, there is still a link with education.
And yes, the researchers did find genetic associations, even a lot of them. In total, the researchers identify 1,260 genetic variants that are significantly associated with personality traits. Of these, 824 had not been found before. Written like this, it all sounds quite spectacular. However, I should note that each individual effect is particularly small! A concrete example to illustrate this: for extraversion, the median effect of a significant variant amounts to approximately 0.009 standard deviations. If two people differed genetically only on such a variant, it would be impossible to tell from their behaviour who should be slightly more or less extraverted as a result. The researchers further estimate that, on average, more than 16,000 independently associated common variants are involved per personality trait.
In other words: there is no extraversion gene, conscientiousness gene, or creativity gene. Personality is genetically extremely fragmented. This is something we have long known for many human traits.
Even when the researchers analyse all the investigated genetic variants together, the effects remain relatively limited. Depending on the personality trait, these genetic variants can collectively explain approximately 5 to 9 per cent of the differences between people. The researchers noted, however, that with better measures, that percentage can be somewhat higher.
That is much less than the 40 to 60 per cent heritability that we often see in twin and family studies. It is not surprising to find such a difference. Twin and family studies attempt to estimate the overall contribution of inherited differences. Meanwhile, this study looks at genetic variants that can actually be measured and linked to personality. Both research methods are also simply different and do not measure exactly the same thing. Also worth repeating is that even high heritability does not mean that the environment is unimportant.
But why do I write a post about this research if everything seems so small? A classic problem in genetic research into behaviour is that parents pass on not only genes but also an environment. Just think of it. Most of the kids grow up with their parents. This may make it hard, for instance, to decide whether a genetic relationship with schooling is truly direct. This may also be somewhat dependent on family, upbringing, social status, partner selection, and more. That is why identical twins raised apart are so attractive for research. The researchers of this study, however, tried to learn more in a different way. They did so through analyses within families.
What did the authors find? When comparing genetic differences within fraternal twin pairs, on average 96 per cent of the association observed in the standard population analysis remains. For educational attainment, that proportion is much lower in previous research, at around 60 per cent. For cognitive ability, it is approximately 80 per cent.
The second analysis, using parent-child data, tells much the same story. When the researchers account for the family environment, about 96 per cent of the genetic association found in the conventional analysis remains. That makes the finding remarkably robust. But be careful with that number: it absolutely does not mean that genes determine 96 per cent of our personality!
Nor does it mean that our genes predetermine our personality. The researchers are very clear about this. Genes influence personality, but they do not fix it. Their effects are probabilistic, and they unfold in constant interaction with our experiences and environment. In our psychology book, we explain that nature and nurture usually interact.
It is also important to note that “generalisable” in this study is not synonymous with “universal”. Most of the data come from populations with European-like genomes and Western contexts. The study also includes African-like samples, but these are much smaller.
They then proceed to relate the genetic associations with all sorts of outcomes, including health, substance abuse, mental disorders, sports, BMI, education, labor market behavior, and even the probability of completing the survey. They also attempt to conduct causal inference on these relationships using Mendelian randomization and identify several potential causal links. However, the authors stress that their findings have to be triangulated with other evidence.
The paper provides evidence refuting two simple stories about personality. The first story is that personality is mainly environmental, if anyone still believes that.
The second is the reverse: that you could read your personality from your DNA. That is clearly incorrect as well. The individual effects are minuscule, thousands of variants play a role simultaneously. Furthermore, even polygenic scores cannot predict an individual’s future behaviour.
So yes: genes matter. Environment matters too. And the most interesting answer remains, as is so often the case, more complicated than the dichotomy between nature and nurture.