In Herrnstein's (1961) experiment, pigeons were placed in a choice situation where they could peck on each of two keys programmed according to independent VI schedules. In a variable interval schedule, a reward (food pellet) is programmed to occur a certain number of seconds after the key has been pecked. The schedule's timer is reset after each reward. As the name suggests, this time interval to the reward is not fixed but varies about a central number of seconds. For instance, in a VI-5 schedule, the number of seconds to each reward (after the triggering peck) would be 5 seconds on average. Hernnstein found that across a variety of pairings of VI-VI schedules, the pigeons' ratio of responses to the two options was equal to the ratio of rewards received from those two options. This relationship is called the matching law.
In this file, we set up a simulation of two VI schedules (all pairs used in Herrnstein's experiment) and provide a model for pigeons pecking the two keys. Model performance is summarized by two numbers: (1) the proportion of responses made to one key (the Left key) and (2) the proportion of rewards associated with that key (the Left key). To the degree that the matching law holds, these two proportions should be similar.