The Long-Disputed Science of Twin Studies

I was surprised to see an article on twin studies by Nancy L. Segal in the July 26^th, 2024 edition of CounterPunch, a politically progressive/left online magazine. Surprised because, while they are often used for other purposes, behavioral twin studies and accompanying heritability estimates are cited by some people in support of biological determinism, IQ hereditarianism, genetic ethnic behavioral and intelligence differences, and in defense of the social and global inequality status quo as a reflection of our supposedly unchangeable biological destiny. Segal participated in the Minnesota Study of Twins Reared Apart (MISTRA), which was financially supported by the Pioneer Fund. The study’s lead investigator, Thomas J. Bouchard, Jr., once concluded, “The similarity we see in personality between biological relatives is almost entirely genetic in origin.” Psychiatry cites twin studies as the gold standard in support of the discipline’s disputed claim that psychiatric conditions such as depression, bipolar disorder, and schizophrenia are genetically based brain disorders only minimally influenced by environmental factors or “triggers.”

On the other hand, due to environmental confounding and other problems, critics sometimes argue that the results of behavioral twin studies should not be interpreted genetically. A confound is an unforeseen or uncontrolled-for factor that threatens the validity of conclusions researchers draw from their data. Confounding occurs when the association between two variables is caused by a third variable influencing both.

Segal is a “go-to” expert enthusiast of twin research, whereas I have published critical analyses of behavioral twin research for over 25 years. Segal did not address the critics’ arguments in her brief CounterPunch article. Portions of my 2015 book The Trouble with Twin Studies attempted to counter claims found in Segal’s 2012 book about the MISTRA, Born Together―Reared Apart: The Landmark Minnesota Twin Study. Segal is a leading behavioral geneticist who argues that twin studies constitute solid science, as do most psychiatry publications. Let’s see if this argument holds up.

Studies of Twins Reared Together (The “Classical Twin Method”)

According to Segal in her CounterPunch article:

“The classic twin study is simple and elegant. Researchers compare the similarities and differences of identical twins to those of fraternal twins. The greater resemblance between identical twins than between fraternal twins is consistent with the view that genetic factors influence the behavioral characteristic or physical trait under study.”

She did not mention that genetic interpretations of behavioral correlation differences between a sample of reared-together identical twins (also known as monozygotic, or MZ) versus same-sex fraternal twins (also known as dizygotic, or DZ) are based on a long-disputed critical assumption known as the “equal environment assumption” (EEA). The EEA holds that identical and fraternal twin pairs grow up experiencing roughly equal environments, and that the only behaviorally relevant factor distinguishing these twin types is their differing degree of genetic relationship to each other (100% identical versus an average 50% fraternal genetic similarity).

However, as most people understand and as research confirms, when compared with same-sex fraternal twins, identical pairs grow up experiencing (1) more similar treatment by parents, teachers, and others; (2) more similar physical and social environments; (3) more similar treatment by society due to sharing a very similar physical appearance; (4) a much greater tendency to spend time together, to have common friends and peer influences, and to model their behavior on each other; and (5) much higher levels of identity confusion and emotional attachment to each other. These findings clearly show that the EEA is false and that the classical twin method not only “overestimates heritability,” but is completely unable to disentangle potential genetic and environmental influences on human behavior.

Nevertheless, although twin researchers usually recognize that identical and fraternal environments are different, they argue that the twin method has been validated by a series of “EEA-test” studies they and their colleagues have performed since the late 1960s. In these studies, researchers “tested” the validity of the EEA in ways other than the only way it can be tested, which is a straightforward determination of whether identical and same-sex fraternal twin pairs grow up experiencing roughly equal environments—yes or no. If the answer is no, as it almost always is, genetic interpretations of identical-fraternal comparisons are not supported. (Readers interested in a complete analysis of the EEA should consult Chapter 4 of my 2023 book Schizophrenia and Genetics: The End of an Illusion.)

The widely recognized much greater environmental similarity experienced by identical (MZ) versus fraternal (DZ) twin pairs indicates that the results of classical twin method studies of behavior are confounded and, therefore, should not be interpreted genetically. Time to rewrite the psychology and psychiatry textbooks.

Studies of Twins Reared Apart (TRA Studies)

Segal also promoted so-called “twins reared apart” (TRA) studies, which she believes “cleanly separate genes and experience.” TRA study researchers attempt to assess genetic and environmental influences on behavioral characteristics studied by psychologists. (Psychiatric twin studies, on the other hand, are based on reared-together twins.) The widely cited and publicized MISTRA is the most well-known TRA study.

There are four basic things people should know about TRA studies: (1) only six have been performed (the first was published in 1937; a later study by British psychologist Cyril Burt is discredited); (2) although researchers studied other areas of behavior, cognitive ability (IQ) has always been the main focus of attention; (3) twins were often volunteers and were not randomly assigned to available homes, and they usually grew up in similar cultural and socioeconomic conditions; and (4) most pairs were not “reared apart” using any reasonable definition of the term.

Concerning the last point, most studied reared-apart identical pairs were only partially reared apart, meaning that a true TRA study has never been performed. Unlike recent studies such as the MISTRA, the authors of the first three studies (published between 1937 and 1965) provided detailed case history information. In these earlier studies, 33% of the twin pairs were separated at 12 months of age or later, 75% had contact with each other while growing up, 23% lived together for at least 12 months after separation or grew up next door to each other, and in 56% of the pairs, one or both twins were placed with a family member. To cite one example, the following passages are found in a case description reported by James Shields in his 1962 TRA study. “Bertram and Christopher” were a 17-year-old male “separated-at-birth” identical twin pair:

“The Mother died the day after the twins were born. The paternal aunts decided to take one twin each and they have brought them up amicably, living next-door to one another. …They are constantly in and out of each other’s houses. …Both twins became garage mechanics. …They have always been closely attached to each other, and now go out courting together. …When they were younger, Christopher used to follow Bertram around ‘as if he were a younger brother.’”

Something to consider: Segal and most other behavioral geneticists count Bertram and Christopher as a pair of “reared-apart identical twins.” Because Segal and her MISTRA colleagues have closely guarded their raw data and information on the twins’ life histories, we can safely assume that, like the other studies, most MISTRA pairs were only partially reared apart.

Cohort Influences. Segal’s and most other behavioral geneticists’ claim that “identical twins reared apart offer pure estimates of genetic influence given that they share all their genes, but not their environments” is false. Even perfectly separated reared-apart twins share a prenatal environment and also share many postnatal non-familial cohort influences due to having been born at the same time and (usually) growing up in the same culture. They often share similar behavior-shaping environmental influences such as extended family, cultural, religious, SES, educational, technological, legal, mass media, regional, political, and oppression/privilege based on ethnicity and gender.

Cohort influences refer to the resemblance in similar-age people’s IQ scores, abilities, behavior, preferences, beliefs, physical condition, and other characteristics caused not by heredity but by experiencing stages of life simultaneously in the same historical period, cultural milieu, and developmental stage. Twins will also elicit more similar treatment from their social environments due to their striking physical resemblance, including physical attractiveness and height. Because in most societies, men and women are treated differently and are socialized from birth to behave, think, and feel in differing gender-specific ways, members of a reared-apart identical pair will behave more similarly for this reason alone.

Identical twin IQ score correlations produced by TRA studies such as the MISTRA are subject to an additional major environmental confound related to the “Flynn effect.” Moral philosopher and IQ researcher James Flynn showed in the 1980s and 1990s that IQ scores worldwide had been increasing by about three points per decade (0.30 points per year), including supposedly “culture-fair” tests such as Raven’s Progressive Matrices (increasing 0.50 points per year worldwide). Reared-apart identical twin pairs found in Segal’s MISTRA and other TRA studies usually grew up in the same country and culture. Therefore, their educational, learning, and cognitive skills development began at the same Flynn effect starting-point in similar era- and location-related “huge IQ advantage or disadvantage” environments, producing yet another non-genetic reason why reared-apart identicals correlate above zero on IQ tests. Contrary to the key MISTRA assumption that reared-apart identical twin behavioral similarity can only be caused by genetics, cohort influences and the Flynn effect indicate that a sizable chunk of these twins’ IQ correlations can be explained not by the genes they share, but by the birth date they share.

Missing Control Group Data. In Born Together―Reared Apart, Segal acknowledged that reared-apart fraternal twins (DZAs) were the official MISTRA control group, designated as such by Bouchard at the beginning of the study in 1979. The MISTRA was the first TRA study to recruit DZAs as controls, which Segal, Bouchard, and colleagues saw as an important improvement over the earlier three studies.

Segal co-authored the famous MISTRA IQ study, which was published in Science in 1990. This article has been cited over 2,400 times. Psychology textbooks usually report its finding of roughly 70% IQ heritability (cognitive ability), even as more recent molecular genetic studies of intelligence and IQ “proxy” measures such as “educational attainment” (EA) have produced dramatically lower heritability estimates and polygenic scores in the 4%-16% range. Apart from the numerous methodological problems with their study and the disputed claim that a heritability statistic estimates the strength of genetic influence on a behavioral characteristic, the MISTRA researchers had to suppress their DZA control group data to find above-zero IQ heritability (explained here and here). No DZA results of any kind were reported in what in Born Together―Reared Apart Segal called the 1990 “Science IQ paper,” which presented “the long-awaited IQ data.”

A colleague and I elaborated on the missing 1990 MISTRA IQ control group data in our recent analysis of the genetic evidence presented in the controversial 1994 book The Bell Curve. To my knowledge, the “long-awaited” MISTRA full-sample DZA IQ correlations remain unpublished to this day, even though the researchers long ago published their full-sample DZA correlations for most non-IQ areas of behavior they studied (e.g., personality).

Cherry-Picked Individual Twin Pairs

A long-running theme of Segal’s writings has been to highlight stories of individual reared-apart identical pairs showing supposedly remarkable behavioral similarities, and she continued this theme in her CounterPunch article. Stories about the supposed “eerie” or “spooky” behavioral similarities in pairs such as the “Jim Twins” have appeared repeatedly in popular books and the mainstream and social media since 1979. Behaviorally dissimilar pairs are rarely discussed in media reports or textbooks.

Far more than human-interest stories, writers and corporate media outlets continue to use these pairs to sell biological determinism to the public in ways that academic research results never could. Although the twins’ stories are often interesting, for reasons that include the cohort effects I described earlier, these selectively reported anecdotal stories provide no evidence that genes influence human behavioral similarities. This conclusion also applies to the famous set of identical triplets whose stories were the subject of the 2018 movie Three Identical Strangers.

The Replication Crisis in Science

The widespread yet misguided acceptance of genetic interpretations of behavioral twin studies reflects serious problems in science as a whole. The scientific research/publication process is experiencing a replication crisis, brought about in part by independent analysts’ 2015 discovery that they could not replicate some key psychology research findings. The original findings were probably non-findings resulting from researchers’ widespread use of “questionable research practices,” sometimes based on vaguely defined characteristics and low-reliability psychiatric diagnoses, including data and conclusions manipulated by researchers behind the scenes to have their manuscripts accepted for publication and to match their own or their funding sources’ expectations and confirmation biases.

Although many researchers did not engage in unethical/questionable practices, the replication crisis is rooted in Segal and Bouchard’s U.S. academic psychology field. For decades, field leaders overlooked and even celebrated shoddy, “p-hacked,” and even fraudulent research based on blatantly false assumptions, data manipulated to match researcher confirmation biases, and financial conflicts of interest. In 1997, the American Psychological Association tried to give a “Gold Medal Award for Lifetime Achievement in the Science of Psychology” to psychologist Raymond Cattell, who in a 1972 book supported the idea of “phasing out” and “ending” what he called genetically “moribund cultures.”

We must now view behavioral and social science research publications through the lens of replication crisis terms, concepts, and perspectives, and the retraction of published scientific research is now common. Top behavioral geneticist Robert Plomin and his colleagues argue that behavioral genetic studies replicate well. However, although this claim has been challenged, the most important question one should ask about a behavioral genetic study is not whether its results replicate but whether its authors’ conclusions were based on sound assumptions.

Conclusion

I have briefly described the fundamental problems with the main twin research methods Nancy Segal promoted in her CounterPunch article. Yet her views are not unusual, as she mainly summarized the current twin research evaluation consensus in psychology, psychiatry, and other behavioral sciences. However, a consensus can be wrong, especially when it is based on highly flawed studies and methods found in politically charged areas such as human behavioral research. Journalist Lawrence Wright observed in 2017, “There’s nothing I can think of in science that is more political than twin studies.” I agree, even as most twin researchers portray themselves as politically disinterested scientists whose data interpretations and conclusions are unaffected by their worldviews and confirmation biases.

Socialist author Phil Gasper described the political thrust of biological determinism, whose foundations are laid by twin research:

“The attempt to explain important features of society in evolutionary or genetic terms—biological determinism—has two goals. First, it tries to convince us that the social order is a consequence of unchanging human biology, so that inequality and injustice cannot be eliminated. Second, in the case of problems that are impossible to ignore, it tells us to look for the solution at the level of the individual and not at the level of social institutions. The problems lie not in the structure of society, but in some of the individuals who make up society. The solution is thus to change—or even eliminate—the individuals, not to challenge existing social structures.”

In contemporary psychiatry, the “solution” is to claim that individuals displaying moods and dysfunctional or socially disapproved behaviors caused by existing social structures and family environments have “heritable illnesses” and “brain disorders” that should be “treated” with powerful drugs.

Segal wrote, “Behavioral and medical science advances have happened largely because twins yield a wealth of information.” I argue instead that behavioral and psychiatric twin studies have yielded mainly genetic misinformation that helps justify political policies, actions, and worldviews antithetical to progressives and the left.

As I showed in my 2004 book The Gene Illusion, twin research was initiated by Francis Galton in the 19^th century. It was further developed by the German, British, and U.S. eugenics movement in the early-to-mid 20^th century. In those days, a twin study was a false-scientific method used to support eugenic theories and practices (including forced sterilization). Today, it is a false-scientific method sometimes used by people promoting behavioral biological determinism, neoliberalism, and the “substantial heritability” of IQ, “personality,” criminal behavior, and psychiatric conditions. A serious and objective evaluation of twin study claims and assumptions in the social and behavioral sciences is long overdue.