In his article "Dynamic Consistency and Non-Expected Utility Models of Choice Under Uncertainty," (Journal of Economic Literature, Dec. 1989), Mark Machina asserted that preference reversal (PR) is caused by a violation of the independence axiom. Amos Tversky, Paul Slovic, and Daniel Kahneman submit, however, that "Observed preference reversal ... cannot be adequately explained by violations of independence...." This paper tests these claims by breaking the independence axiom into its two component parts: mixture and replacement separability.
Part one of this paper will give some background of the preference reversal field and identify the core questions that this paper will try to answer. Part two will detail experimental design. Part three will present the results and draw conclusions from them. Part four will summarize the most important findings of the paper.