Deviations from optimal choice : skilled performance, feedback, and Bayesian decision making

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Richard Alexander Bauman (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site:
Richard L. Shull

Abstract: In the present study, reinforcement initiated the probabilistic assignment of food to either the right or the left lever and each assigned food was held available. But, if food were assigned to the right lever, a 5-sec minimum waiting time (a DRL 5 sees schedule) was activated and the temporal availability of food assigned to this lever was limited to 5 sees. If food were assigned to the left lever, a longer (L sees) DRL schedule was activated. Since a press on the left lever could not satisfy the 5-sec DRL requirement, the optimal form of choice for the present procedure was alternation, if and only if the temporal accuracy of a rat were perfect. Only if this were true, would the absence of food following the first leverpress perfectly predict the availability of food for switching. Each of 5 rats was presented with each of the six possible combinations formed from two magnitudes of L (10 and 40 sees) and three magnitudes of its probability (0.50, 0.75, and 0.95).

Additional Information

Language: English
Date: 1982
Bayesian statistical decision theory
Decision making in animals
Rats $x Psychology

Email this document to