Fourth, given the relative rarity of our outcome measures, we had an insufficient number of DZ twins to obtain useful estimates of key parameters. Therefore, we added to the sample near-aged sibling pairs who have the same genetic relationship to one another as do DZ twins and, like DZ twins, closely share their rearing environment. There is a possibility that resemblance in these pairs might be lower for than for DZ pairs (because their level of environmental sharing may be slightly less—they do not share the same womb at the same time) although that is not supported by a comparison of the results available to use (see Tables 2b and 2c). If this is the case, then this would result in some over-estimate of genetic effects and a likely underestimate of shared environmental effects. A major bias in parameter estimates is unlikely because when we examined all forms of DA in both twins and all full siblings, we estimated the “special twin environment” – which would cause excess resemblance DZ pairs compared to all siblings – to account for only
