The role of the response-reinforcer relation in delay-of-reinforcement effects

The role of the response-reinforcer relation in maintaining operant behavio r under conditions of delayed reinforcement was investigated by using a two -operandum (i.e., two-key) procedure with pigeons. Responding on one key wa s reinforced under a tandem variable-interval differential-reinforcement-of -other-behavior (tandem VI DRO) schedule. The schedule defined a resetting unsignaled delay-of-reinforcement procedure in that a response was required when the interfood interval of the VI schedule lapsed, but further respond ing during the DRO component on either key reset the time interval. This en sured a fixed delay duration between any response and reinforcement. Respon ding on another key, physically identical to the first one except for spati al location, otherwise was without consequence. The location of the key cor related with the delay-of-reinforcement procedure varied between sessions a ccording to a semirandom sequence. Differences in response rates between th e two keys were greater, with proportionally higher rates on the key correl ated with the delay-of-reinforcement procedure, the longer the delay-of-rei nforcement procedure remained correlated with the same key. Differences in responding on the two keys also increased within individual sessions. These results suggest that the response-reinforcer relation is the primary deter minant of responding when responding is acquired and maintained with delaye d reinforcement.