Professor Crow agrees "for the most part with Jensen's analysis." He does suggest qualifications when drawing inferences from existing studies in biometrical genetics. First, he notes his reservations about the reality of the mathematical assumptions implicit in analysis of variance models. Second, he draws attention to the limited sample size available in studies of twins and siblings reared apart and asks how representative such groups are. Third, he notes that predictive models have inherent limits when new, qualitatively different, treatments are introduced into the environment.

This content is only available as a PDF.
You do not currently have access to this content.