Changing Choices: Why and How, If At All

Abstract

The phenomenon treated in this chapter is well-known: People sometimes do not do what they apparently have decided to do. For example, Beach, Campbell, and Townes (1979) developed and tested a decision aid for counseling couples who were not sure whether they wanted to have a (another) child. In a study with 200 couples the subjectively expected utilities (SEUs) of both options were calculated, however, the couples were not informed about them. Predictions based on the SEUs were compared with the couples' actual behavior after two years, that is, whether they had a child, were expecting one, or had at least tried to conceive. The predictions were very good for those couples whose SEU against having a child was higher than the SEU in favor of having a child. However, for those couples whose SEU in favor of having a child was higher than the SEU against having a child, the predictions were relatively poor; there were many "false alarms," that is, these couples had not tried to have a child despite their positive attitude. Whether the basis of their original preference was fragile, or whether it was changed due to later events, was not directly controlled. The data revealed, however, that the differences between the SEUs in favor of having a child and the SEUs against having a child were much higher for the hit group than for the false-alarm group. Perhaps the latter group was less sure about their preferences and therefore hesitated to implement its decision.