Rapid demand curves for behavioral economics

A method for determining the relationship between food consumption and the price of food (demand function) in behavioral economic experiments is described. Although previous methods have generally required as long as 40 days, the present method can generate a complete demand function within 7 days, and therefore may be more suitable for use in the evaluation of drugs, toxins, and physiological/anatomical interventions. Moreover, measures of circadian rhythmicity, post-reinforcement pause durations, and interresponse times can also be generated. Three experiments tested the stability of the method in a variety of procedural manipulations: repeated exposure to the procedure, increasing versus random daily food price, and size of daily changes in food price. The procedure generated demand functions that were similar to those that require more extended testing, and the demand functions were not generally affected by procedural manipulations. Body weight, which can also affect consumption, generally decreases with increases in the price of food; so this variable should be recorded and used as a covariate in analyzing demand functions. With the exception of circadian rhythmicity, the other measures were stable across procedural variations and showed expected changes as a function of food price: postreinforcement pause durations increased as price increased, but interresponse times did not.