Intertrial interval, concurrent and previous trial reward magnitude, and pattern responding in the discrete trials lever press apparatus

The present study examined the effect of intertrial interval (ITI), prior trial reward magnitude, and concurrent trial reward magnitude on responding in the discrete trials lever press apparatus. Two groups of 10 rat Ss each were given 11 trials per day of single alternation (SA) partial reinforcement (PRF) and 3 groups were given 11 trials per day of SA varied reinforcement (VRF) for 60 acquisition days. These trials were given under a long ITI (60 sec.). A transfer phase followed in which one half of the Ss in each group were shifted to a short ITI (30 sec.). The ITI was unchanged for the other half of the Ss and all other conditions were identical to acquisition. A subsequent transfer returned all Ss to the conditions of acquisition. The PRF Ss, but not the VRF Ss, reached a significant level of pattern responding by day 60. Transfer Increased pattern responding in all groups; all but one VRF group reached a significant level of pattern responding during transfer. Responding in the PRF groups was a function of both the prior trial and the concurrent trial reward magnitude, while responding in the VRF groups was a function of only prior trial reward magnitude. Responding in the second transfer in all groups returned to the terminal acquisition level, it was concluded that conditioning occurs in the PRF groups, but not the VRF groups. Responding, including the pattern responding observed in transfer, in the VRF groups was a function of an unconditioned response decrement.