The science in the selection of these variables is far from precise at this time. Similarly, in the siblings reared apart cross-fostering design, it is important to consider potential “third” variables. The many environmental differences between adoptees and their biological siblings pose an unavoidable third variable threat. Further, because couples who adopt are screened by agencies for financial security, but no such screening is applied to biological parents who chose to parent, there are unavoidable economic and other differences between birth parents and adoptive parents that make a full cross-fostering design unattainable. Finally, in the IVF design, prenatal effects cannot be disentangled from postnatal effects without inclusion of the surrogacy group, which was small in the Cardiff IVF study (n = 23). In addition, unlike the parent-child adoption design and the cross-fostering design, the IVF design does not include measurement of characteristics of the donor parent (i.e., the genetically-related parent). Thus, the designs described in this article are quasi-experimental, naturalistic approaches that are subject to a lack of precision, but that allow for novel insights into the role of the environment in influencing children’s risk for psychopathology that could not be discerned without genetically-sensitive methods.
