Superstitious key pecking after three peck-produced reinforcements.

Researchers gave pigeons three chances to peck a key for grain. After that, grain arrived no matter what they did.

The birds kept pecking for 50 sessions even though the key no longer controlled food.

Only three accidental reinforcers were enough to create long-lasting superstitious behavior.

The pigeons acted as if their pecks still caused the grain to appear.

Catania et al. (1982) extends this idea. They showed superstition grows when five different colored lights appear. Birds pecked at different speeds under each color even though food rates stayed the same.

Gibbon (1967) and Neuringer (1969) came first. These studies proved pigeons will peck keys on fixed-interval schedules. The 1970 paper builds on that baseline by showing pecks can persist without any real schedule at all.

Falcomata et al. (2012) used a similar long-session design. They tracked how pause and peck patterns stabilize over time, giving us a yardstick for how odd the 1970 superstition effect really is.

Your client may repeat tiny, useless behaviors if you accidentally reinforce them even once. Check that every reward you deliver is truly tied to the target skill. If you see mystery responses during probe sessions, stop and reset contingencies before the superstition locks in.