Measurement of individual preferences. The importance of "situation-specific" variables

Women in two Ontario cities were asked to express their preferences about a medical decision: A fictitious patient with cancer was to be treated either conservatively without hope of cure, or radically, by a risky treatment having cure or immediate death as possible outcomes. Situation-specific variables were varied systematically across the sample by varying the levels of the fictitious patient's physical and psychosocial health and achievement motivation. The results obtained indicate that individual preferences shift in response to changes in the situation in which the decision is made. These findings raise questions about the uses to be made of measurements of preferences, especially in relation to what is meant by a "best decision."