Reinforcement delay: some effects on behavioral contrast.

Henton (1972) tested pigeons in a two-part schedule. Part one gave grain right after each peck. Part two delayed grain by several seconds or removed it.

The birds stayed in the same box and the key color told them which part was active. Researchers watched if pecking sped up in the unchanged first part.

Contrast still appeared even when the delay did not cut the number of reinforcers or the birds’ peck rate in the changed part.

Longer delays and full extinction made the effect stronger, but the numbers across birds were not statistically different.

Ginsburg et al. (1971) showed that longer extinction before food gives bigger contrast. Henton (1972) adds that simply stretching the delay can do the same.

Thomas et al. (1974) later proved that just lowering reinforcement frequency, while keeping response rate flat, is enough to drive contrast. Together the three papers show contrast can pop up without any drop in how often the animal responds.

Rogers-Warren et al. (1976) looks like a contradiction: they got zero contrast when reinforcement rates were equal across parts. The key difference is control. A held frequency exactly the same; W let frequency drift while adding delay. The studies together warn us: watch both time and rate when you program schedules.

You now know contrast can surprise you even when the client’s response rate or your reinforcer count looks stable. If you add a wait time before delivery, or stretch the schedule, watch for a burst in the other activity. Plan extra prompts or differential reinforcement to keep the spike within safe limits.