NONOPTIMAL AND COUNTERINTUITIVE CHOICE

University Of California San Diego, La Jolla CA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (adapted from applicant's abstract): The PI proposes to investigate the determinants of decision-making in contexts wherein the decisions are apparently non-optimal and/or counterintuitive. By better appreciating the variables controlling decisions that we make in these apparently anomalous cases, we should be in stronger position to enhance our general appreciation of the determinants of choice. The first set of proposed experiments continues the investigators' recent research with both pigeon and human subjects on the role of unconditional probabilities of reinforcement (base-rates) on choice. This research suggests that human tend to ignore these probabilities in certain behavioral settings resulting in non-optimal choices, while pigeons choose optimally in comparable settings. They propose experiments that should further clarify the variables that determines whether or not humans and pigeons incorporate base-rate information in their decisions. The second set of proposed experiments' addresses the important issue of how different types of problem-solving experiences, including instructions, affect the efficiency and accuracy of subsequent problem solving. The task is the classic problem solving task involving simple computation, developed originally by the Luchins: Will subjects who have been instructed to use rules develop more rigid (and subsequently inefficient) problem-solving techniques than subjects who have acquired the rules on their own or than subjects who have been exposed to a series of novel problems? To enhance the generality of the conclusions the investigators also propose to assess the effects of such training in the acquisition and transfer of behavior and novel-conditional matching-to-sample problem set. Finally they propose to assess the effects of experience, including instructions, on decisions made in problems involving the conjunction fallacy. This latter subset of studies has the additional goal of assessing the role of feedback in repeated conjunction-fallacy problems. The final set of proposed experiments explores in humans a strikingly counterintuitive finding. Over a broad range, choice of pigeons for two outcomes in the concurrent chains procedure remains invariant over changes in the duration of the outcomes as long as a constant difference is maintained between them. This counterintuitive finding is predicted by delay-reduction theory developed in the PI's laboratory over the past thirty years. Most people's intuition and most theories of choice require a different pattern of choices, namely converging preferences with increasing durations (and constant difference). The investigators propose to replicate the constant-difference invariance experiment with human subjects and with four types of choice procedures in order to excess its generality. If the invariance holds with humans, it would place a strict constraint on theories of decision-making. All of these proposed experiments should enhance our appreciation of variables affecting choice