Dissociation of value and response strength.

Researchers let pigeons choose between two keys. Each key paid off on a variable-interval schedule.

The birds first showed which key they liked better. Then the team watched if the birds kept picking that key even when it paid less.

Sometimes the pigeons stayed on the key that gave fewer treats. Their early "value" did not match their later "response strength."

The result weakens simple maximization theory: animals do not always pick the richest schedule.

Henson et al. (1979) saw the opposite. Their pigeons always moved to the higher payoff. The difference is timing: E’s birds could switch any moment, while W’s birds faced stricter change-over delays that locked them in.

Aman et al. (1993) later repeated the idea with rats. Long lever runs paid less, yet the rats kept running. Both studies show local contingencies can override overall rate.

Peters et al. (2013) extended the point to chain schedules. Pigeons pecked more in a middle link when its stimulus promised better odds, even though total payoff stayed the same. Value and strength split again.

If clients stick with a response that earns fewer reinforcers, check the local contingencies, not just the overall rate. Thin schedules, change-over delays, or conditioned stimuli can glue behavior in place. Loosen those local constraints before you raise the payoff elsewhere.