Abstract

The principal goal of psychophysics is to describe the functions that transform the objective variables into their subjective equivalents. The matching law has been used to describe the function translating the objective strength of a rewarding stimulus (e.g., the concentration of a sucrose solution) into its subjective impact. In contrast, the psychophysics of the cost of a reward has been little studied. This is a salient lacuna. The present experiments examine opportunity cost, the price paid for a reward in terms of the time taken away from competing activities, such as procurement of alternate rewards, resting, grooming, and exploration. The reward is electrical stimulation delivered to the medial forebrain bundle. It is the subjective interpretation of these values (reward strength and cost) that an animal uses for goal selection. A widely held assumption is that the animal computes the payoff of a goal (an index of how worthwhile it is for it to choose a goal) as the ratio of the subjective reward strength to the subjective cost required to obtain the goal. This assumption is the basis for the present experiments: changes in subjective reward strength (manipulated by the frequency of stimulation) are used to compensate for changes in opportunity costs by the animal, and changes in opportunity costs are used to compensate for changes in reward strength. The present experiments estimate the function that transforms objective opportunity costs into subjective opportunity costs under the assumed definition of payoff. In Experiment I, the change in frequency required to offset a constant proportional change in price is determined, providing an estimate of the first derivative of the subjective-price function. This experiment demonstrates that as the time intervals are shortened, subjective opportunity cost levels off and deviates substantially from the objective cost. In Experiment II, the change in price required to offset a fixed difference in frequency is determined which provides an estimate of the subjective-price function itself. This experiment demonstrates that subjective costs approximate objective ones when the time intervals involved are relatively long. The two experiments complemented each other: the first revealed the scalar range of the objective-subjective cost relationship while the second demonstrated where this relationship breaks down. It has been implicitly assumed that the relation between objective and subjective costs approximate each other; these experiments showed that this is true but that the relationship breaks down at low costs. The function mapping objective costs into subjective ones (the "subjective-cost function") would prove useful because if the cost of a reward were to be used as a parameter that is manipulated in other experiments, it is important to choose a value of cost that the animal will accurately interpret