Is There Something More Important behind Framing?

Abstract

Medical students as well as staff members at a gas company were asked to choose between alternative rescue programs for the wounded in a gas explosion. The decision task was presented in both positive and negative versions. Within each of these versions were two pairs of choice problems in which the second was derived from the first by the substitution of a common constituent from each alternative. The conditions were designed so that the results should be predictable from the postulates of either Savage′s (1954) sure thing principle or Tversky and Kahneman′s (1981) framing effect. However, what is found is that Savage′s sure thing principle holds in positive frames but fails in negative ones whereas Tversky and Kahneman′s framing effect holds in the first pair of choices but fails in the second. The present data could not be fully explained by these postulates and are taken as stressing the need for an alternative explanatory mechanism.