Thomas J. Bouchard, Jr., David T. Lykken, Matthew McGue, Nancy L. Segal, Auke Tellegen
6 min
This landmark study, known as the Minnesota Study of Twins Reared Apart, investigated the relative contributions of genetics and environment to human psychological development. By assessing over 100 sets of monozygotic (identical) and dizygotic (fraternal) twins who were separated in infancy and raised in different households, the researchers sought to disentangle the effects of nature and nurture on intelligence, personality, and social attitudes.
Participants underwent an intensive, week-long assessment involving approximately 50 hours of medical and psychological testing. The researchers utilized multiple, independent instruments for each domain—including the Wechsler Adult Intelligence Scale (WAIS) for IQ, various personality inventories, and occupational interest surveys. To ensure the validity of their findings, the team also conducted a systematic assessment of the twins' rearing environments to determine if similarities in adoptive homes could account for the observed behavioral correlations.
The results revealed a striking degree of similarity between monozygotic twins reared apart, often rivaling the similarity seen in identical twins reared together. Specifically, the study estimated that genetic factors account for approximately 70% of the variance in IQ. Furthermore, the researchers found that the influence of the shared family environment on adult personality and intelligence is surprisingly minimal. Instead, they propose that genetic factors may influence psychological development indirectly by shaping the environments that individuals seek out, create, or elicit from others.
These findings challenge traditional psychological theories that emphasize the primary role of the home environment in shaping individual differences. By demonstrating that genetic factors play a significant role in a wide array of traits—from cognitive ability to occupational interests—the study underscores the importance of biological predispositions. However, the authors caution that these results do not diminish the value of education or parenting, but rather suggest that interventions may be most effective when tailored to an individual's specific, genetically influenced talents and inclinations.
Alex: [probing] And what about prenatal environment? Monozygotic twins share a womb. Couldn't shared prenatal exposure be doing some of the work?
Sam: [grounded] The authors address this directly. Their argument is that prenatal history rarely drives long-term psychological similarity — and if anything, prenatal complications like vascular steal tend to decrease within-pair similarity rather than inflate it. So the prenatal confound, if anything, biases against finding high correlations. The observed similarity survives that correction.
Alex: [slight head-tilt] What about post-reunion contact? These twins eventually find each other. Couldn't shared adult experience be creating the similarity rather than reflecting it?
Sam: [steady] The researchers tested this explicitly. Contact time accounts for essentially none of the IQ similarity. And the directionality argument runs the other way — it's the pre-existing similarity that predicts increased contact, not contact producing convergence. That's an important distinction for causal inference here.
Alex: [checking understanding] So the main finding holds against those alternative explanations. But there's an age effect in the MZT data worth unpacking — the correlation for twins reared together tends to decline as they get older. What's driving that? [[RP_SECTION:developmental-trajectories-and-age|Developmental Trajectories and Age]]
Sam: [thoughtful] That's one of the more theoretically interesting patterns in the dataset. As twins age out of the shared household and accumulate their own non-shared experiences — different careers, relationships, social contexts — their cognitive profiles diverge slightly. When you compare adult MZA correlations to older MZT cohorts, the gap between them narrows considerably. It suggests the shared rearing environment has a real but transient effect: it shapes the developmental trajectory early on, but it doesn't determine the endpoint. The genome appears to reassert itself over time.
Alex: [analytical edge] Which raises the generalizability question. This sample is predominantly middle-class and Western. Can these heritability estimates travel to contexts of genuine deprivation? [[RP_SECTION:generalizability-and-context|Generalizability and Context]]
Sam: [direct] The authors are explicit that they can't. Heritability is a population statistic — it's a function of how much environmental variance exists in the sample. In a relatively homogeneous, resource-adequate environment, genetic differences explain most of the remaining variance in outcomes. But compress the genetic variance or expand the environmental variance — as you would in conditions of severe deprivation — and the estimate shifts. High heritability in a Minnesota sample doesn't mean environment is irrelevant; it means environment was relatively uniform in that sample.
Alex: [measured] So the 70% figure is partly a reflection of the cultural and economic context of the study, not just the genome.
Sam: [calm] Exactly. Heritability isn't a fixed biological constant. It's a ratio that changes with the range of environments you're sampling from. The Minnesota findings are internally valid and methodologically careful — but extrapolating the estimates to populations facing extreme resource constraints would be a misuse of the design. The study tells us a great deal about how genetic architecture operates within the envelope of modern Western societies. What happens outside that envelope is a different question, and one this dataset can't answer.
Alex: [reflective] That's a useful boundary to draw. The design is strong within its scope — the MZA comparison is genuinely informative — but the scope itself is the constraint. Thanks for listening to ResearchPod.