Date of Award:


Document Type:


Degree Name:

Master of Science (MS)



Committee Chair(s)

Timothy A. Shahan


Timothy A. Shahan


Amy L. Odum


Gregory J. Madden


Resistance to extinction of single-schedule performance is negatively related to the reinforcer rate that an organism experienced in the pre-extinction context. This finding opposes the predications of behavioral momentum theory, which states that resistance to change, in general, is positively related to reinforcer rates. The quantitative model of extinction provided by behavioral momentum theory can describe resistance to extinction following single schedules in a post-hoc fashion, and only if the parameters of the model are allowed to vary considerably from those typically derived from multiple- schedule preparations. An application of the principles of Bayesian inference offers an alternative account of extinction performance following single schedules. According to the Bayesian change-detection algorithm, the temporal intervals of non-reinforcement that an organism experiences during extinction are compared to the temporal distribution of reinforcers that the organism experienced during baseline. A transition to extinction is more readily detectable when the previously collected distribution of reinforcers in time
is populated with relatively short intervals (i.e., when more frequent reinforcement was experienced during baseline). The Bayesian change-detection algorithm also suggests that changes in reinforcer rates are more detectable when organisms have temporally proximal experience with frequently changing rates. The current experiment investigated this novel prediction. Pigeons pecked keys for food under schedules of reinforcement that arranged either relatively dynamic reinforcer rates or relatively static rates across conditions. Following each period of reinforcement, resistance to extinction was assessed. Persistence was greater following static contingencies than following dynamic contingencies for the majority of subjects. These data provide support for the Bayesian approach to understanding operant extinction and might serve to extend behavioral momentum theory by offering change detection as an additional mechanism through which extinction occurs.



Included in

Psychology Commons