Effects of varying delay of reinforcement and postreinforcement intervals on learning of retardates

All combinations of two levels of delay of reinforcement (0 and 15 seconds) and two levels of postreinforcement intervals (1 and 16 seconds) were used in a two choice discrimination learning task with institutionalized retardates. Significantly more correct responses were given under 0-second delay. Latency of responding was significantly shorter under 0-second delay than under 15-second delay, and was also shorter under 1-second postreinforcement than under 16-second postreinforcement conditions. The interaction of postreinforcement interval by blocks of trials produced a significantly greater shortening of latency in the 1-second condition over the five blocks per day. The latency results indicated that delay interval should not be confounded with postreinforcement interval since both intervals influence latency of responding.