Reaction latency (StR) as a function of the number of reinforcements (N).

Abstract

59 albino rats, 12 to 14 weeks of age at the beginning of training, were taught to push a short horizontal bar to the left for a reward of a small cylinder of relished food automatically delivered at once following the response. Trials were made once every 24 hrs., after a 22-hour no-feeding interval. Training was continued until reaction latencies, graphically recorded, had ceased to decrease. For the median animal this plateau came at about the 60th trial. Both the reaction latencies of individual distributions were markedly skewed, their modes falling near the short latency end. The median latencies (str) of the group as a whole (with data equalized for individual differences in rate of learning) were found to have the functional relationship to the number of reinforced training trials (N) which is well fitted by the equation, stR = 33.0 N-1.2 -0.25.