The Effect of Framing on Choice

Abstract

Tversky and Kahneman suggested that choice between a "sure thing" and a risky option of equal expected value is affected by option phrasing. When options for fighting an "Asian disease" were framed positively, 72% chose the sure thing; but when options were framed negatively, only 22% did. Decision problems like the Asian disease one, however, have produced varied results. Experiment I explored three reasons for the varied results: differences between the samples in risk-taking propensity, differences in gender composition, and use of different decision problems. A significant interaction of framing, sex, and problem was observed. Women were affected by framing, while men generally were not. Further, the size and direction of framing effects varied across the decision problems. Experiment 2 investigated risk-taking propensity and field independence as mediators of the sex by framing interaction. Only framing was significant. Neither field independence nor risk-taking propensity interacted with framing.