An analysis of progressively delayed reinforcement.

A procedure based loosely on Ferster\u27s 1953 study was developed to examine the effects of progressively delayed reinforcement on keypecking in pigeons. The main procedure combined progressive schedule components with delayed reinforcement so that delays separating responses from food delivery were incremented after each successive reinforcer. Three experiments were conducted to describe behavior under these progressive contingencies. Experiment 1 examined the effects of imposing progressive delays on variable-interval and fixed-interval schedules while controlling for the simultaneous decreases in reinforcer rate. Progressive delay sessions alternated daily with interval sessions of immediate reinforcement constructed from the interreinforcement intervals generated by the immediately preceding delay session. Behavior adjusted to the increasing delays within single sessions in a way similar to behavior under traditional delay manipulations which occur across conditions. In Experiment 2, breaking points (the last delay before responding ceases for 7 min) were similar across four delay step sizes. In Experiment 3, breaking points were longer and response rates were higher when progressively delayed reinforcers were signaled. Results are discussed in terms of progressive contingencies and delayed reinforcement