Abstract
Eight groups of 8 hooded rats were assigned to a 2 × 2 × 2 design of 32 or 128 acquisition trials, 50% or 100% reward, and chains of 1 or 8 bar-presses. Starting speeds were examined, as were trials to an extinction criterion of 5 min. without completing a response chain. During acquisition, performance was an increasing function of the number of trials and percentage reward. For the 128-trial groups, at 1:1 reward 50% rewarded Ss responded significantly faster than 100% Ss; however, at 8:1 reward there was no significant difference due to percentage reward. The number of trials to extinction was an increasing function of the number of acquisition trials and a decreasing function of length of response chain. While all other groups reflected a rise in performance with an increase in number of acquisition trials, the 100%-8:1 groups showed no change in performance due to this variable.