MacIntosh Study Faulted

December 5, 1994

Houston MacIntosh's “Attitudes and Experiences of Psychoanalysts in Analyzing Homosexual Patients” discussed the results of a thirteen-item questionnaire administered to a convenience sample of 285 psychoanalysts. Subjects reported their analytic experiences with 1,215 homosexual patients. MacIntosh's study has important limitations, particularly in the area of research design.

Our first concern is that the study confounds political and scientific issues. Participants were recruited with a letter suggesting that their responses could help counter ideas fostered by Richard Isay that most psychoanalysts are poorly educated about homosexuality and negatively biased toward homosexual patients. This personalized appeal must have affected who chose to participate in the study and who declined.

A second problem is MacIntosh's failure to define what he meant by homosexual orientation. There is considerable disagreement in our field about how patients' sexual orientation should be classified, especially when they have experienced overt bisexual fantasy or engaged in bisexual activity. This problem in research design could have been avoided had homosexual orientation been defined in the questionnaire.

Third, MacIntosh states that “23 percent of homosexual patients in analysis change to heterosexuality” (p. 1203). Since this is based on the treating analysts' memory and judgment, with no systematic follow-up, it cannot be considered genuine observation. To reliably assess the rate of conversion would require reports from third-party evaluators and/or the patients themselves at an interval after the analyses had ended. Without this, the patients' desire to please the analyst, the analysts' incomplete information, and failure to assess behavioral change as lasting over time render these data pseudofactual. MacIntosh suggests that his data replicate findings reported in the well-known Bieber study. However, Bieber et al. did not assess sexual orientation change on f