Tellegen · Journal of personality and social psychology 1988 · cross-sectional twin study · n=402 twin pairs (804 individuals)

Personality similarity in twins reared apart and together.

Cited 1339 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional twin study comparing reared-together and reared-apart pairs (by design analogy)

PubMed 3397862 · doi:10.1037//0022-3514.54.6.1031 · record verified 2026-08-29

What was done

Researchers administered the Multidimensional Personality Questionnaire (MPQ; 11 primary scales, 3 higher-order scales) to adult twin pairs across four groups: 217 monozygotic reared-together, 114 dizygotic reared-together, 44 monozygotic reared-apart, and 27 dizygotic reared-apart pairs (total 402 pairs). Using maximum-likelihood techniques, they fitted a full four-parameter biometric model (additive genetic, nonadditive genetic, shared family environment, and unshared environment) alongside five reduced models to evaluate the contributions of genes and environment.

What was found

Purely environmental models failed to fit any of the 14 MPQ scales. The full model provided a satisfactory fit for all scales, yielding heritability estimates ranging from .39 to .58. The contribution of shared family environment was negligible for 12 of the 14 measures, showing a meaningful effect in only 2 scales. In addition, 3 measures showed evidence of significant nonadditive genetic effects.

Why it matters

This landmark reared-apart twin study provided empirical evidence that adult personality traits are substantially heritable (roughly 40% to 60%) and that growing up in the same household accounts for very little similarity in adult personality.

Limits

The sample of reared-apart pairs was relatively small (44 MZ and 27 DZ pairs). Personality was assessed exclusively via self-report on a single instrument (the MPQ). The findings depend on the assumptions of biometric structural equation modeling, and the abstract does not report specific demographic characteristics or confidence intervals for the heritability parameters.

Cited by