Choice and the relative immediacy of reinforcement.

Kydd et al. (1982) let pigeons peck two keys. Each key led to grain after a delay.

The twist: both keys gave the same average wait time. One key just split the wait into a short pause then a longer pause. The other key did the reverse.

Birds chose hundreds of times. Researchers asked: will the birds favor the key that gives the first pause sooner?

The pigeons showed no preference. They pecked each key about equally.

Even though one key produced grain faster on half the trials, the birds did not care. Total wait time, not the speed of the first bit, controlled choice.

Steege et al. (1989) later added humans to the same setup. People, unlike pigeons, did favor the option that raised their overall payoff. The bird data stayed the same, showing the null finding is reliable for pigeons.

Singh et al. (1982) looked similar but kept total wait unequal. When they made one side truly faster, pigeons switched. This supports R’s point: birds watch total delay, not early tokens.

Kim et al. (2025) reviewed decades of choice studies. They list Kydd et al. (1982) as a caution: giving a tiny early reward does not help if the whole schedule stays long.

When you build token boards or delay schedules, remember that clients sum the whole wait. Adding a quick point or sticker early will not boost motivation unless you also shorten total time to the big reinforcer. Check the entire chain, not just the first link.