Has a New Twin Study Meta-Analysis Finally “Settled” the Nature-Nurture Debate?

Jay Joseph, Psy.D.

In the May 20th, 2015 edition of the Huffington Post, Macrina Cooper-White reported on a recently published meta-analysis (analysis of combined studies) of 2,748 twin studies performed in 39 different countries published between 1958 and 2012, which looked at more than 17,000 physical, medical, and psychological characteristics (traits). Cooper-White concluded that the results of this study indicate that the nature-nurture debate “may now be over,” as the study found that nature (genetic) and nurture (environmental) influences “are virtually tied.”

In the original study published online in the prestigious scientific journal Nature Genetics, Tinca Polderman and a group of prominent colleagues concluded that their “results provide compelling evidence that all human traits are heritable.” This conclusion is not new, however, as behavioral geneticists have been making this claim for decades. Polderman and colleagues believed that their meta-analysis of twin method comparisons of reared-together MZ and DZ twin pairs “provides the most comprehensive analysis of the causes of individual differences in human traits,” which “will guide future gene-mapping efforts.”1 (The researchers excluded so-called reared-apart twin pairs from their study.)

Amazingly, the researchers failed to mention that genetic interpretations of MZ-DZ correlational differences are based on the validity of the twin method’s controversial “equal environment assumption” (EEA). The EEA requires twin researchers to assume that reared-together MZ and DZ pairs grow up experiencing roughly equal environments, and that the only factor distinguishing these two types of twin pairs is their differing degree of genetic relationship to each other (100% MZ versus an average 50% DZ).  This assumption is obviously false, as the evidence clearly shows that MZ pairs grow up experiencing much more similar environments and treatment than do DZ pairs, and that MZs experience much greater levels of identity confusion and psychological closeness as well.2

Because twin method MZ-DZ comparisons are based on the false assumption that MZ and DZ pairs experience equal environments, it doesn’t matter whether researchers pool together the results of 5 twin studies, 500 twin studies, 2,748 twin studies, or a million twin studies. Like the individual studies, the pooled results for behavioral characteristics can be completely explained by the non-genetic (environmental) influences experienced to a much greater degree by MZ versus DZ twin pairs. Two wrongs don’t make a right, and 2,748 environmentally confounded twin studies pooled together don’t make a genetic finding, at least as it relates to human behavioral differences.3

The classic textbook example of the folly of concluding that correlation equals cause is the finding that ice cream sales correlate with the violent crime rate. Both increase as the weather gets hotter in the summer months, and it obviously would be wrong for an investigator to conclude that eating ice cream causes people to commit violent crime—even if the same results, and accompanying conclusions, are found in 2,700+ pooled studies. The authors of the Nature Genetics study have committed a similar correlational folly by concluding that the greater behavioral resemblance of MZ versus DZ twin pairs is caused by genetic factors, and as usual the popular press follows right along without mentioning that critics have thoroughly exposed this folly in numerous publications going back to the 1930s.4

Regardless of the (predictable) pooled results and conclusions found in the Polderman et al. Nature Genetics study, because the EEA is false, genetic interpretations of all past, present, and future MZ–DZ twin method comparisons in psychiatry and in the social and behavioral sciences in general must be rejected outright. The classical twin method as a measure of “heritability” and of genetic influences on psychiatric disorders and other behavioral characteristics remains one of the great pseudoscientific methods of our time, and will eventually be added to the list of discarded pseudosciences where we now find alchemy, craniometry, and mesmerism.

However, because false genetic theories and techniques are needed by powerful economic and political interests, it is unlikely that the twin method will be abandoned in the absence of significant social and political change.

* * * * *


1. Polderman, T. J. C., Benyamin, B., de Leeuw, C. A., Sullivan, P. F., van Bochoven, A., Visscher, P. M., & Posthuma, D., (2015), Meta-Analysis of the Heritability of Human Traits Based on Fifty Years of Twin Studies, Nature Genetics, published online 5/18/2015. doi:10.1038/ng.3285

2. Joseph, J., (2004), The Gene Illusion: Genetic Research in Psychiatry and Psychology under the Microscope, New York: Algora; Joseph, J., (2015), The Trouble with Twin Studies: A Reassessment of Twin Research in the Social and Behavioral Sciences, New York: Routledge.

3. A “confound” is an unforeseen or uncontrolled-for factor that threatens the validity of conclusions researchers draw from their studies. Confounding occurs when the association between two variables is caused by a third variable that influences both. In the twin method, the more similar environments experienced by MZ twin pairs confound genetic interpretations of MZ-DZ correlational differences.

4. See Joseph, 2004, Chapter 2.

Support MIA

Enjoyed what you just read? Consider a donation to help us continue to produce content, provide up-to-date research news, offer continuing education courses, and continue building a community for exploring alternatives to the current paradigm of mental health. All donations are tax deductible.

Select Payment Method
Personal Info

Credit Card Info
This is a secure SSL encrypted payment.

Donation Total: $20.00


  1. Behavioral genetics is founded on an absurd premise; the social construction of science is evident in the acceptance of the ridiculous concept of the EEA. The harmful concept of “mental disorders” is widely considered to be based on science; thank you for exposing the pseudo-scientific foundation of popular psychology theory.

    Best wishes, Steve

  2. The conclusions that nature and nurture are tied seem common sense regardless of the validity of any study denying or confirming this conclusion.

    Furthermore, is it this study that confirms nature and nurture are tied or is this a fact that can be deduced by observing human behaviors in different contexts?

    Even though this new study seems accurate, like all psychological studies, it relies upon many assumptions that have never been teased out for their accuracy or scientific validity. For example, what concretely and verifiably comprises ‘psychological’ characteristics, or traits?Psychological characteristics are a mysteriously undefined, vague, and not at all uniformly understood set of ‘things’ for lack of a better word because there are no tangible words that clearly define what a psychological characteristic actually is. Most psychological definitions are tautological. A ‘psychological’ trait is caused by a person’s psychology.

    Do alleged psychological characterisitcs reside inside the brain and make the brain think certain thoughts? Do psychological characteristics reside smack dab inside the genes and do these genes cause people to think and behave as they do? Or do psychological characteristics swim around in the cells and inform bodies how to behave? Or is it all three? Or more likely, is it none of the above?

    I don’t think there is any such animal as a psychological characteristic. Our brain stores memories. Our brain assesses and organizes information in order to form cogntiive conclusions about internal and external streams of information it continuously receives. When the brain receives information to organize it gives itself lots of feedback in order to cue the relevant parts of the brain that have the proper mechanisms to deal with it. Once the brain has ciruclated a stream of information to all the relevant areas in itself, it forms a predictive decision for what to do next.

    The brain is very specific in how it takes in, orders, concludes, and then formulates predictive decisions for what to do next. The brain is an organ that manages information. It is not an organ stuffed with psychological characteristics.

    A person’s sensory, motor, and nervous systems will determine how he is able to make sense of information and then manage the outcomes of that information, not his so called psychological characteristics. A person is a decision in a moment, not a result of his or her psychological characteristics.

    My point being, this study sounds logical as do many psychological studies, but to untangle itself from all the contraversy and contradictions, psychiatry has to start from scratch. It’s foundational assumptions are just that, assumptions. The assumptions have to be cleared away so a scientifically accurate theory of the brain can inform it’s obervations, research, and interpretations.

    • Steve, I feel this position (the idea that nature / nurture have equal influences – and also the idea that the degree of influence of nature/nurture can be quantified) is fundamentally mistaken. I encourage people to check out Evelyn Fox Keller’s book, The Mirage of a Space Between Nature and Nurture.
      After reading this I got a new conception of how genes and environment work. They interact complexly and dynamically, and rely on each other in such an interpenetrating way that to try to tease them apart and quantify how much each contributes to development is misguided and fundamentally misunderstands how change/growth happens.

      • Interesting comments. Your comments speak to my own ideas about why behavior modification is so confusing and damaging to children. Each child is making sense of information in his or her environment in ways that make sense to their uniquely calibrated sensory, motor, and nervous systems. To intervene in how they evaluate, decide, and respond to the world around them so it makes more sense to a standard of behavior is what I believe to be one of the main causes of mental illness. The book you referenced sounds interesting as well.

  3. I have long been bothered about the notion of us being mere genetic robots. It didn’t seem to fit what I observe in the everyday world. For years I have agreed that genetics may play a part in temperament, but a small one. This new meta analysis study had me delving into articles, and I stumbled upon the Minnesota Twin study. With the Minnesota twin study, I too was victim of the shock and awe of the raised apart twins that were reunited after several years. The minute details (wearing rubber bands on wrists, flushing toilets before and after, same name of wives and kids) stunned me into thinking maybe heredity plays 100%. I mean, how can you account for all those similarities? I later found a summary of your first book chapter by chapter, which helped address some of my own laymen concerns.

    I do appreciate your cohort argument, and I had not realized the “behind the scenes” work with the photographs. I have been wondering: they highlighted 5-7 or so identical twins reared apart that had amazing similarities, but don’t say anything about the other pairs. Were any completely different in personality in regards to observation and the test? I forget how many they had in the study, (MZRA) but it seemed quite a few. Do you have insight into those?

    I agree that it seems money tends to flow toward genetic studies, which in turn leads to biased results. Money does not flow towards family/environment studies.

    Just for fun, with all scientific data aside and just from your own informed opinion, what would you conjecture to be the genetic component of personality?

  4. Why is it that you present EEA as an either/or, valid/invalid issue that makes twin studies valid or invalid? I’ve read the EEA criticisms of twin studies, including Burt and Simons that only seems to reference you, but I’ve seen nothing in them that actually invalidates twin studies. What these criticism do is put error bars on the estimates of genetic, shared environment, and unshared environment effects. That is, the degree to which the numerical estimates output from the model change due to the degree of inaccuracy of any of the inputs, including EEA, is a matter for error propagation studies. These studies clearly show the output estimates are relatively insensitive to deviations from the EEA. The error propagation of EEA has also been studied for decades and the minimal effects on the outcomes has been known, e.g., as outline by Barnes et al.

    Furthermore, these criticisms via EEA only affect the model that provides the numerical estimates of percentages of genetic and environmental (shared and unshared) components; it does not at all affect the science of behaviours as resulting from genes, nor replaces that with any sort of environmental explanation for the data.

    For example, the EEA does not affect the high correlation of behaviours with gene-relatedness nor low correlation with shared environment. With or without the numerical model you still have the problem of explaining such data. That is, given the results of behavioural and personality tests, you would generally not be able to tell much better than guessing which MZ twins were raised in the same household and which were not, which DZ twins were raised in the same household and which were not, and which unrelated peers were raised in the same household (adoptees) and which were not, but you would be able to tell which sets of results came from MZ twins (highly correlated), DZ twins (~half as correlates), and unrelated (not correlated). The explanation would also have cover why these correlations between pairs of people also correlate with their genetic relatedness. To reiterate, none of that data relies on the EEA. The EEA only shows up in the model for estimating percentage influences, which again is quite robust to EEA deviation.

    So if you have some references that justify both your rejection of the model (despite error propagation studies showing robustness) and your rejection of behavioural genetic data in general, please provide them. That would truly be scientifically important if such science exists to reject these things, yet all you refer to is the EEA assumption which doesn’t reject either the model or the underlying data.

    It appears to me, reviewing all of your work, you are beating a dead horse. The science is clear, but you seem stuck on an either/or mentality of either a perfect model or else it is zero value, which of course is not valid reasoning.