Response rate, reinforcement frequency, and behavioral contrast.

Thomas et al. (1974) worked with pigeons on a two-part schedule. In Part A the birds got food on a VI 5-min schedule. In Part B the food stayed the same timing but the amount dropped. The team kept the birds' peck rates steady by hand. They wanted to know if less food alone would make the birds peck faster in Part A.

This setup tests pure reinforcement frequency. Response rate never changed in Part B. Any jump in Part A pecks would show frequency, not rate, drives contrast.

When food frequency fell in Part B, pecking rose in Part A. The birds worked harder for the same VI 5-min food. Contrast appeared even though their own rate stayed flat.

The study proved reinforcement frequency alone can push contrast. You do not need big response drops in the other part.

Rogers-Warren et al. (1976) seems to disagree. They equated food across parts and saw no contrast. The key gap is equality. V lowered food counts; A kept them the same. Both papers agree: contrast needs a real food difference, not just schedule names.

Henton (1972) backs the idea. He held food rate steady but added delay. Contrast still showed up. Together these studies tell us any drop in food value—less food, slower food, or delayed food—can spark contrast.

Reynolds et al. (1968) went further. They gave extra stimulus feedback and saw contrast while food actually rose. Contrast can happen even when the other part earns more, as long as the animal notices a change in payoff.

For BCBAs this means watch your reinforcer counts, not just the schedule name. If you thin reinforcement in one activity, clients may work harder in the next, even if their behavior looks the same. Check rates across settings after any reduction program. You might see bursts you did not expect.