Environmental mediation and the twin design

Behavior genetic twin designs are increasingly used to study the effects of a measured environment whilst controlling for genetic variation. In this research note, we show that, in the context of the classical twin design, (1) when the environmental variable is necessarily shared between twins, the notion of controlling for genetic influence is logically flawed and (2) when the environmental variable varies between twins in the same family, partial control for genetic influence is possible, but only if appropriate analytic models are used, which is commonly not the case. Based on a simple simulation study, recommendations are given as to which methods should be applied and which should be avoided.