Date of Graduation


Document Type


Degree Type



Eberly College of Arts and Sciences



Committee Chair

Kennon A Lattal

Committee Co-Chair

Dwight Harshbarger

Committee Member

Michael Perone

Committee Member

Claire St Peter

Committee Member

Cole Vonder Haar


Roughly 50 years of research has demonstrated that choice changes as a function of the amount of reinforcement available from two concurrently available schedules of reinforcement, and this function has been labeled the generalized matching law. The matching law, however, historically only accounts for choice between two variable-interval schedules of reinforcement. If the matching law is to be a general perspective for how organisms allocate behavior amongst several alternatives, then it ought to account for behavior when reinforcers depend on work requirements in addition to time requirements. To study these contexts, the present experiments employed concurrent ratio schedules that were programmed nonindependently. As such, responses to one ratio schedule simultaneously incremented the response counter for both schedules. The probability of reinforcement increased on both schedules as responses were allocated to either schedule. In Experiment 1, response allocation of pigeons was assessed when three lengths of changeover delays (0 s, 2.5 s, and 10 s) were in effect. Experiment 2 compared choice in concurrent variable-ratio and fixed-ratio schedules. Experiment 3 controlled the number of reinforcers delivered to two concurrently-available variable-ratio schedules. When concurrent variable-ratio schedules were in effect, the matching law described choice well in all three experiments. Choice on concurrent fixed-ratio schedules was inconsistent across pigeons. The main variable controlling response allocation in all experiments was the distribution of reinforcement. By controlling the number of reinforcers delivered to each schedule, choice can be brought under precise control even when ratio schedules are in effect. These results support extensions of the generalized matching law to contexts in which reinforcers are contingent on response requirements rather than time requirements.