That makes a lot more sense, Wolfgang. IF you're looking for the possible genetic components of any trait or behavior, twin studies are a great place to start. Of course this does NOT mean that twins are more likely to display the trait or behavior - just that twins are ideal subjects for this type of study. This article from the Economist about a recent twin study in London supports what you've said ...

...behavioural traits with a large genetic component are more likely to be shared by identical twins than fraternal twins. Conversely, those traits with a large environmental component will be shared by identical and fraternal twins in equal measure. Applying appropriate statistical techniques to the actual amount of shared behaviour observed allows the relative contributions of genes and environment to be worked out....

Based on the teachers' assessments, the researchers identified the naughtiest 10% of the individuals in their sample—in other words those with severe conduct disorder. They then subdivided these children into those with psychopathic traits and those without and asked, in each case, whether an individual's twin showed bad behaviour, psychopathy, or both.

Their analysis showed that bad behaviour without psychopathy has relatively little genetic component—less than a third. By contrast, four-fifths of the difference in behaviour between the general population and children with psychopathic traits seems to lie in the genes.

Hmmmm... that last paragraph is a bit of a head-scratcher though...