The recurrence of negatively reinforced responding of humans.

Alessandri et al. (2015) asked adults to press a key to turn off loud noise. The task removed an annoying sound, so the behavior was kept going by negative reinforcement.

After the key stopped working, the team watched for three comeback patterns: resurgence, renewal, and reinstatement. They wanted to see if these relapse effects happen when the original reward is escape from something bad.

Resurgence and renewal showed up almost every time. When the new response no longer worked, people went back to the old key press that used to shut off the noise.

Reinstatement was shakier. Sometimes the noise alone brought the key press back, sometimes it did not. The study proved that negatively reinforced behavior can spring back after extinction.

Wilson et al. (2016) seems to disagree at first glance. They found almost zero resurgence when kids had to walk farther to place a ball. The trick is response effort: harder work wiped out the effect that Jérôme saw. Same relapse process, but effort is the off-switch.

Laureano et al. (2023) turn the 2015 finding into a spreadsheet. Their RaC2 model lets you type in your session data and predict the exact minute problem behavior will pop back up. The lab demo is now a planning tool.

Novak et al. (2020) repeat the renewal piece in a fake office. Desirable work behavior also returned when the pay rules changed, showing the effect is not tied to annoying noise or key presses.

If you extinguish escape-maintained problem behavior, plan for it to return. Use the RaC2 sheet from Laureano et al. (2023) to pick the day you will add booster sessions. If the response is easy, raise the effort like Wilson et al. (2016) did—move materials, add steps, or close a lid. Expect renewal whenever you shift contexts, and pre-teach the new routine the way Novak et al. (2020) did with office tasks.