THE "ARTIE" SIMULATION OF OPERANT CONDITIONING

Abstract

Computational modeling is emerging as a new method for researchers to test theories of behavior. Computational models that incorporate adaptive networks are especially attractive to behavior analysts because of the selectionist nature of these models. A novel adaptive network model of operant conditioning, called "Artie," is presented. Results from simulated laboratory experiments with Artie replicated laboratory results from real animal s, lending support to selectionist accounts of operant conditioning. In general, adaptive network models of behavior are probably better suited for testing theories of behavior than for teaching us about the functioning of the brain.