Matching and maximizing in a self-control paradigm using human subjects

Abstract

Five experiments examined adult human females' sensitivity to variation in amount and delay of reinforcement, as well as their impulsiveness (preference for smaller, less delayed reinforcers over larger, more delayed reinforcers), self-control (the opposite of impulsiveness), and indifference (equal preference for larger, more delayed reinforcers and for smaller, less delayed reinforcers). Each experiment used four to five subjects, concurrent variable-interval schedules, reinforcers that consisted of opportunities to earn points exchangeable for money, and varied relative reinforcer amount and delay separately and/or together. Four experiments used a deduction procedure: a random number of points was repeatedly removed from the display counter at regular intervals, independently of the subjects' behavior. Three experiments used independent concurrent schedules; the other two used nonindependent concurrent schedules. One experiment allowed the subjects to play a radio during laboratory sessions. In all experiments the subjects' behavior was better predicted by a molar maximization model than by the generalized matching law. However, the indifference that occurred when the ratio was used did not closely approximate molar maximization. The deduction procedure appeared to have no effect on the subjects' choices. Use of the nonindependent concurrent schedules seemed to result in the subjects showing indifference and low sensitivity to changes in relative reinforcer amount and delay. Use of the radio also seemed to result in indifference. Unlike the behavior of pigeons, the behavior of adult human subjects within a self-control paradigm appears to be well described by molar maximization, but it may be possible to disrupt such behavior by environmental stimuli.