Quantitative studies of reinforcement relativity.

Kazdin (1977) tested whether the matching law works when rats choose between two very different rewards. The rats could run in a wheel or drink sugar water. The team changed how often each reward followed a response. They recorded where the rats spent their time across three experiments.

The rats’ time split matched the payoff split. If wheel running paid off twice as often, the rats spent about twice as long on the wheel. The simple matching equation predicted the results within a few percent. Motivation shifts made the fit a little loose, but the rule still held.

Rilling et al. (1969) showed the same time-payoff match in pigeons eight years earlier. Kazdin (1977) proves the rule crosses species and reward types—birds with seed to rats with exercise.

Deluty et al. (1978) ran the next study in the same lab. They swapped sugar for mild shock and found the matching law still works for punishment. Together the three papers show one equation handles both good and bad outcomes.

Kydd et al. (1982) later added a twist: changeover behavior follows its own invariant rule. That refinement keeps the core matching idea but sharpens the math for rapid switches.

You now have evidence that the matching law guides choices even when the options feel nothing alike—say, screen time versus snacks. When a client splits time oddly, check the payoff rates first. A quick tally of reinforcers earned on each activity often explains the ‘bias’ and shows where to intervene.