Percentage reinforcement and choice.

E and his team worked with pigeons in a small lab chamber. Two keys lit up. Pecking either key could pay off with grain.

The twist: one key paid every time (100 %). The other paid only 30 % or 50 % of the time. The birds could switch keys whenever they wanted.

Sessions ran daily. The scientists counted which key the bird chose and how much grain it actually got.

Birds landed on the side that gave more grain per minute. They did this even when that side was the 'unreliable' 30 % key.

Schedule 'honesty' did not matter. Only the obtained pay-off rate guided choice.

Shimp et al. (1974) showed pigeons love signalled reinforcement. E’s birds also picked the richer side, but they did not need a signal to do it. Together the papers say: signals help, yet rate still rules.

Dukhayyil et al. (1973) and Leung et al. (2014) moved the same rule to kids. Children, like pigeons, gravitate toward the schedule that delivers the most goodies. The mechanism crosses species.

Deluty (1976) flipped the coin and used punishment. Higher shock rate on one key pushed birds to the other key, again showing choice tracks consequence rate, just in the opposite direction.

When you set up concurrent programs, stop worrying about looking 'fair' or 'predictable.' Count how many reinforcers the client actually earns per minute on each option. Put more payoff on the behavior you want to grow. The learner’s choices will follow the numbers, not the schedule label.