Utility, induced utilities, and small worlds.

Abstract : Decision theory in its present state of development fails to provide sufficient criteria for the unique formulation of a decision problem. The resulting ambiguity may lead to quite disparate interpretations of the same decision task by different individuals. The inconsistencies found in the measurement of the ''utilities'' of objects which in reality serve as means to an end and in the Chipman-Ellsberg-Fellner paradox may arise from differences between experimenter and subject in their formulation of the decision task. This ambiguity can be reduced either by analyzing the decision task in the largest context feasible or by introducing structural constraints on the formulation of the decision problem. (Author)