CHOICE, DYNAMIC REINFORCEMENT, AND CHANGEOVER RESPONSE REQUIREMENT

Recent studies have shown that choice does not require steady state performance to show sensitivity toreinforcement. This idea was further explored in a choice situation in which a changeover responserequirement was manipulated. The parameters of the generalized matching law (Baum, 1974) were usedto assess changes in preference due to an increasing fixed-ratio changeover-response requirement. Theratio of reinforcement provided by two levers (27:1, 9:1, 3:1, 1:1, 1:3, 1:9, 1:27) was manipulated betweensessions to prevent steady state concurrent performance. In different conditions, the changeoverrequirement took on values of 1, 4, 8, 16 and 32 responses. The results showed local effects of reinforcementon choice behavior; from one reinforcer to the next, the distribution of responses favored thealternative associated with the highest probability of reinforcement. Although sensitivity to reinforcementdid not increase with increasing changeover response requirements, as the rats experienced thedynamic reinforcing environment along the experiment the value of the parameter s stabilized above 1.0.