Subjective framing and attitudes towards risk

Abstract Numerous examples of framing effects have been reported since Kahneman and Tversky's (1979) original article on prospect theory. However, the focus has been on imposed frames. We design an experiment to elicit subjective frames and to verify if these frames yield the same systematically different choices observed in experiments using imposed frames. We then compare the predictive power of frames with that of knowledge of risk preferences. For our experiment, we find no relationship between choice and attitudes towards risk, while a clear relationship exists between choice and frame.