Concurrent Schedule Performance with Young Infants
Bonem, Elliott J.
MetadataShow full item record
The problem: In operant research on infant conditioning, extinction sessions often are conducted in demonstrating schedule control over the response. However, extinction sessions are often accompanied by disruptive behaviors which terminate experimental sessions. Concurrent reinforcement schedules provide a methodology for demonstrating schedule control over the response without relying on extinction sessions. The purpose of this study is to determine whether infant responding can be acquired, maintained, and systematically controlled by concurrent reinforcement schedule contingencies. Procedure. During daily sessions, three young infants were exposed to several concurrent variable-interval schedules. A mobile rotation, contingent on a foot kick, was occasionally available on each schedule component. Head turns served as the changeover response. Relative time allocation was assessed as was matching between the logarithm of time (and response) ratios and log-reinforcement ratios. Findings. Without relying on extinction sessions, schedule control was demonstrated for all infants. The proportion of time spent in a given schedule component systematically varied as the programmed availability of contingent mobile movement was varied. By tending to allocate responses and time equally between the two schedule components, irrespective of programmed reinforcement, all infants undermatched and showed a positive bias for the richer schedule component. Conclusion. Concurrent reinforcement schedules provide a methodology for demonstrating systematic control over infant responding. Although the matching relationship seems to generalize to young infants, the variables which are the source of undermatching and bias have not been identified. Recommendations. To gain better experimental control over infant responding, the sources of bias and undermatching must be identified. For example, future research could determine whether bias results from using qualitatively different reinforcers, or whether better matching results from incorporating a changeover delay.
35 leaves. Advisor: Larry A. Alferink.