Repeated, within‐session resurgence

Cook et al. (2019) ran a three-step loop inside one lab session. First, they reinforced one response. Next, they paid for a new response and stopped paying for the first. Finally, they removed all pay to see if the first response popped back up.

They repeated this loop many times in the same sitting. The goal was to make resurgence happen fast and often so researchers could study it quicker.

The old response did surge back almost every loop. How high and how long it surged changed between subjects.

Higher pay during the first step made the surge bigger. Higher pay for the new response made the surge smaller. These patterns matched earlier work that used days-long phases.

Friis (1998) also flipped pay rates within one session. That paper looked at how hard it is to stop the response, not at resurgence. Both show that within-session swaps give clean data, so you can run many mini-experiments in one sitting.

Lejuez et al. (2001) doubled the pay rate and saw double the resistance to disruption. Cook saw a similar pay-size effect, but for resurgence height. Together they say reinforcer rate controls both how hard the response dies and how easily it returns.

Cao et al. (2026) split persistence into five parts in mice. Cook’s short loops could let us test those parts one by one in humans without waiting weeks.

If you study relapse, you can now get ten resurgence probes in one hour instead of ten days. Try the three-step loop next time you test why a child returns to old problem behavior. Watch the first response come back each time you remove pay. Note how big and how long it lasts. You will have a full dose-response curve before snack time.