Persistence during and resurgence following noncontingent reinforcement implemented with and without extinction

Saini’s team compared two ways to run noncontingent reinforcement. One group got free reinforcers while problem behavior was ignored. The other group got the same free reinforcers but problem behavior was also put on extinction.

Each adult participant first earned candy for pressing a button. Next, the candy stopped and a brief NCR phase began. Finally, all reinforcement paused to see if the old button pressing would surge back.

NCR without extinction cut button pressing faster at first. When the break phase arrived, however, that group showed a bigger resurgence spike.

Adding extinction to NCR slowed the initial drop but produced a smaller, shorter comeback later. The pattern held for most, not all, participants.

Fisher et al. (2018) later showed the same trade-off in children. They built momentum by packing reinforcement before thinning, echoing Saini’s finding that early density matters for later resurgence.

McIntyre et al. (2002) had already proven that rich histories make behavior tougher to extinguish. Saini’s data extend that rule to resurgence after NCR.

Slocum et al. (2025) seem to disagree at first glance. They found that skipping extinction at the start gave faster suppression in 3- to young learners. The difference is timing: Slocum measured only the first ten sessions, while Saini tracked the later resurgence window. Fast early win, bigger later loss.

If you need rapid calm for safety, NCR alone can work short-term. For lasting reduction, pair NCR with extinction from day one. Expect a slower start but steadier long-term gains and less comeback during future transitions or caregiver mistakes.