Objective Benefit Versus Subjective Perception in the Theory of Risk-Sensitive Foraging

Abstract

The organism''s responses to variability in resource distributions is a compound effect of both the perceived (subjective) probability of the occurrence of different reward levels or benefits and the utility that results from such benefits. To date, research has focused primarily on the utility function that assigns values to different outcomes of acts. Some kinds of observed behavior (like the preference of uncertain rewards over certain rewards with equal expectation) have led behavioral ecologists to postulate that organismal utility functions are both convex and concave but over different portions of the range of rewards. Though this is a plausible and likely explanation, observed patterns of behavior and their tests cannot distinguish the effects of shifting subjective probabilities at different levels of reward expectation. If subjective probabilities are skewed such that rare favorable events are perceived to be more likely to occur than they objectively are, then organisms may show apparently risk-seeking behavior without implying a change in the curvature of the underlying utility function. Shifts in probabilities of this sort do not guarantee the concavity of utility, nor does the existence of such shifts exclude the possibility that utility is both convex and concave. The two problems are separate, but only one (the shape of utility) has received attention. I present a graphical method for partitioning these effects on the basis of the concept of "acceptance sets". The method leads to testable prediction that relate easily to the experimental systems commonly employed in the analysis of risk-sensitive foraging.