Controlled vs. Naturalistic Experiments: Application of the Life Table Method

Abstract

I would like to discuss this subject by drawing from my own experiences, which may present a familiar topic from a different vantage point. I’l1 describe research on the outpatient treatment of primary depression, in which I’ve collaborated since 1967 with Drs. Gerald Klerman and Alberto DiMascio, and Ms. Brigitte Prusoff. This research is relevant to the topic since it included a twenty-month study with both controlled and naturalistic phases. I will also draw upon my training in epidemiology, where procedures for the study of cholera or the statistical techniques used for obtaining mortality data can be useful models for psychiatric research. This discussion will illustrate that the limitations and advantages of naturalistic and controlled experiments are not that clear. Moreover, the statistical technique of life tables can be used to approximate quasiexperimental conditions for the analysis of data collected naturalistically.

Appreciation is expressed to Gerald L. Klerman, M.D., Alberto DiMascio, Ph.D., and Brigitte Prusoff, M.P.H., for the use of collaborative data and to Gerald L. Klerman, M.D. for comments on the material presented in this paper. This research was supported by U.S.P.H.S. Grants MH 13738, MH 15650, and MH 17728 from the Psychopharmacology Research Branch, National Institute of Mental Health, Department of Health, Education, and Welfare.