Predicting the Value of Untried Options

University Of Oklahoma Norman Campus, Norman OK

Investigators

Abstract

Choice behavior is usually studied by pitting known alternatives against each other; however, in nature new opportunities frequently appear. Animals face the far more difficult problem of deciding between options of known vs. unknown value. This study examines whether animals solve such problems by using a set of potential rules tied to regularities in the environment. The study is carried out in a naturalistic foraging setting, where laboratory animals must choose between novel and familiar patches of buried food. Conditions are arranged such that certain patterns of choices should emerge in some contexts, and quite different patterns in other contexts, if appropriate rules are triggered. Broadly, the project is aimed at determining if natural selection has shaped the decision-making processes of animals such that difficulties posed by uncertainty are minimized. These decision processes are expected to reside in relatively simple cognitive mechanisms that cross species boundaries. Advances in understanding choice behavior have wide applicability. Beyond furthering the understanding of animal behavior, the rules uncovered in this project may be useful to social scientists trying to predict consumer-buying patterns and to engineers designing adaptive systems.