Abstract

Behavioral economists have proposed that human preferences are constructed during their elicitation and are thus influenced by the elicitation procedure. For example,
different preferences are expressed when options are encountered one at a time or concurrently. This phenomenon has been attributed to differences in the “evaluability” of
a particular attribute when comparison to an option with a different value of this attribute is or is not available. Research on the preferences of laboratory animals has often been carried out by means of operant conditioning methods. Formal treatments of operant behavior relate preferences to
variables such as the strength and cost of reward but do not
address the evaluability of these variables. Two experiments assessed the impact of procedural factors
likely to alter the evaluability of an opportunity cost (“price”): the work time required for a rat to earn a train of rewarding electrical brain stimulation. The results support the notion that comparison between recently encountered prices is necessary to render the price variable highly evaluable. When price is held constant over many trials and test sessions, the evaluability of this variable appears to decline. Implications are discussed for the design of procedures for estimating subjective reward
strengths and costs in operant conditioning experiments aimed at characterizing, identifying and understanding neural circuitry underlying evaluation and choice.