Resistance To Change As A Function Of Stimulus-reinforcer And Location-reinforcer Contingencies.

The team worked with pigeons on a two-key setup. Each key showed a different color. One color came with more food than the other.

They then moved the food machine. It could drop grain near either key. The birds still had to peck the same color to earn food.

Next they stopped all food and watched which key the birds kept pecking. They wanted to know if the birds followed the color cue or the place that once had more food.

The birds stuck with the color that had paid off most, even after the food machine moved. The key location did not matter.

Resistance to extinction followed the stimulus-reinforcer link, not the total food each spot had delivered.

The finding backs up Matousek et al. (1992) and Prasher et al. (1995). Those labs also saw that richer components fight disruption harder, but they used extinction or free food as the test.

Cullinan et al. (2001) looks like a clash. They saw VR schedules give high response rates yet low resistance. The key difference is they held reinforcer rate steady while changing how responses produced it. The 1996 study held the response rule steady and only moved food around, so the cue-value link stayed clear.

Costa et al. (2024) later repeated the same rule with humans and point-loss. Higher-rate components again held strong, showing the color-food bond works across species and disruptors.

When you program reinforcement, think about the cue, not just the amount. A token board, picture card, or S-delta signals value to the learner. If you shift materials around the room but keep the same cue, the behavior should stay steady. If you change the cue itself, even rich schedules may lose their punch. Pick clear, consistent signals and protect them during transitions.