chapter  5
The Minnesota Study of Twins Reared Apart I
Biases, Assumptions, and Other Problem Areas
Pages 26

This chapter examines the major problems and biases in the MISTRA, many of which were also in evidence in the classical studies. In most MISTRA publications reported by Segal, the researchers concluded in favor of genetics on the basis of twin data used in their model-fitting procedures. Kamin's analysis was not expected to support the researchers' positions, so he was not granted access to the raw data. A finding that MZA and DZA sample correlations do not differ argues against the genetic position because MZA pairs' more similar genetic relationship versus DZA pairs did not lead to their greater behavioral resemblance, as predicted by genetic theories. The Minnesota researchers attempted to adjust their data to eliminate age and sex effects. Beckwith and colleagues requested much more information about the twins' life circumstances and degree of separation, Bouchard and colleagues answered this request with a correlation coefficient.