Abstract
This paper assesses the impact that the death of two training analysts had on their analysands. The dependent variable, which is length of analytic training, was chosen because it is objective, easily measured and should reflect the disruption due to an analyst's death. Patients of the deceased analysts had a significantly increased length of training (Fisher's exact probability, p < .02). This work reviews a number of relevant methodological issues and attempts to show how one can examine observable phenomena empirically. It is hoped that clinicians will begin to address how their psychotherapeutic work could be assessed using scientific methodology.
