The decision-making under risk and uncertainty literature is built upon two canonical models---Expected Utility Theory (EU) for risk, and Subjective Expected Utility Theory (SEU) for uncertainty. Since their inception, researchers have identified a number of paradoxes that represent normatively appealing violations of these theories. Most notably, Allais' Common Consequence Problem demonstrates a frequent violation of EU in favor of certain outcomes, while Ellsberg's Three Color Problem demonstrates a violation of SEU by avoiding bets that require constructing subjective beliefs over uncertain outcomes. Both of these paradoxes have spurred the literature into analyzing not if, but why, these violations occur. As a consequence, increasingly more complex models have been developed to motivate these seemingly paradoxical behaviors. This thesis experimentally explores these motivations, focusing specifically on the two paradoxes formerly mentioned, and the space of novel problems between them. Finally, in addition to the canonical problems, we explore a more modern phenomenon violating EU. This phenomenon is referred to as "mixture effects", which occurs when an individual makes different choices after facing the same set of options multiple times.
Jack Field Adeney (Mon,) studied this question.