Successive reversals of a classically conditioned heart-rate discrimination.

Neuringer et al. (1968) trained rats on a heart-rate discrimination. One tone meant "speed up," another meant "slow down."

After the rats met criterion, the team flipped the rules. The old "speed up" tone now meant "slow down." They repeated this reversal several times.

Each reversal took fewer sessions than the one before. Early reversals were slow; later ones were fast.

The learning curve looked just like curves seen in lever-press or key-peck reversal studies.

Rees et al. (1967) watched monkeys learn a visual switch task. The monkeys moved through clear stages: guess, stick, then truly discriminate. Neuringer et al. (1968) add the reversal twist and show the same stage-like speed-up.

Mansell et al. (2002) later copied the reversal logic with humans who have severe ID. They added a 5-second delay before the correct stimulus. The delay worked like the later reversals here—accuracy jumped quickly.

Millard (1979) kept the reversal format but let one pigeon’s pecking serve as the cue for another pigeon. Both birds still showed faster reversal learning across rounds, proving the effect holds even when the cue is social.

If your learner stalls after the first reversal, don’t panic. The 1968 rats tell us the second, third, and fourth flips get easier. Keep the procedure steady; the speed-up will come. You can also borrow the 2002 delay trick or the 1979 social-cue idea to nudge the process along.