Response-reinforcer relations and the maintenance of behavior.

Researchers worked with three pigeons in a small chamber. Each bird could peck a lit key for food.

They compared two setups. In one, every peck produced grain after a short delay. In the other, grain dropped on a timer no matter what the bird did.

Birds pecked far less when food came for free. Even a delayed peck-produced pellet kept more behavior alive than the timer did.

As the delay after a peck grew from 2 to 8 seconds, pecking still stayed above the free-food level.

Davis et al. (1972) already showed that longer signaled delays cut VI key pecking. S et al. now add that any link, even a slow one, beats no link at all.

Van Hanegem et al. (2014) later saw the same pattern in children. Contingent praise for compliance beat handing out the same reinforcers on a time clock, echoing the pigeon result across species and settings.

Barnard et al. (1977) had found that autoshaped pecks vanish when the response-food gap tops four seconds. The 1987 data line up: once delays stretch past that mark, rates fall, but they still stay above response-independent levels.

When you shape new skills, keep the contingency crystal clear. If you must add a delay, insert a bridge or a signal so the learner still sees that their response “caused” the reinforcer. Avoid non-contingent token drops or edible giveaways unless the goal is to reduce motivation; free items can accidentally suppress the very behavior you want to strengthen.