ABA Fundamentals

Resurgence in a discrete‐trial procedure in rats

Fujimaki et al. (2025) · Journal of the Experimental Analysis of Behavior 2025
★ The Verdict

Resurgence appears even when only one response per trial is allowed, so plan for it in discrete-trial teaching.

✓ Read this if BCBAs who run discrete-trial training or extinction procedures in clinics or schools.
✗ Skip if Practitioners who only use free-operant or naturalistic setups and never run extinction.

01Research in Context

01

What this study did

Fujimaki and team worked with lab rats in a small box. Each rat could press one of two levers. Pressing the first lever used to bring food, then it stopped. The rats learned to press a second lever for food instead.

Next the researchers stopped food for the second lever. They let each rat make only one press per trial. They counted how often the rats went back to the first, no-longer-paid lever. That return is called resurgence.

02

What they found

Every rat showed strong resurgence. When the new lever no longer paid off, the rats returned to pressing the first lever even though it still never gave food. The effect showed up early and stayed through the whole session.

Resurgence did not need free-access conditions. One press per trial was enough for the old response to bounce back.

03

How this fits with other research

Fuhrman et al. (2016) showed that careful schedule thinning during functional communication training can stop later resurgence in children. Fujimaki’s rats prove the basic effect still lurks if you simply halt reinforcement, giving a reason to use Fuhrman’s thinning tactic.

Hatton et al. (2005) also worked with rats and found that drugs can speed extinction after fixed-interval training. Fujimaki adds that, even without drugs, the extinguished response is ready to reappear when the alternative fails.

Barszcz et al. (2021) and Saini et al. (2015) cut stereotypy with response-interruption and redirection, another extinction-based method. Fujimaki’s clean discrete-trial set-up gives future researchers a simple way to test ways to stop resurgence before trying them with kids.

04

Why it matters

You now have lab proof that resurgence can hide inside tight, one-response-per-trial programs. If you run discrete-trial teaching, know that simply stopping reinforcement for a new skill can bring back the old problem behavior. Build in gradual thinning or alternate reinforcers before you pull the plug, just as Fuhrman et al. did with FCT.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Before you stop reinforcing a replacement behavior, thin the schedule over at least three sessions and watch for the old response to pop back.

02At a glance

Intervention
extinction
Design
single case other
Population
not specified
Finding
positive
Magnitude
large

03Original abstract

Resurgence is a transient recovery of a previously extinguished target response following a worsening of reinforcement conditions for an alternative response. Laboratory studies with nonhuman animals typically assess resurgence in free-operant situations where subjects can freely emit responses. The purpose of the current study was to investigate whether resurgence would be observed in a discrete-trial procedure where only a single response could occur in each trial, using rats as subjects. The experiment consisted of three phases, and each session ended after 200 trials. All trials began with the insertion of target and alternative levers and ended once a response was emitted. In Phase 1, both target and alternative responses were reinforced with a probability of .25. In Phase 2, the target response was extinguished while still reinforcing the alternative response with a probability of .25. Finally, resurgence was tested by placing the alternative response on extinction. All rats showed robust resurgence in this highly constrained discrete-trial situation. We also found that the latencies of resurged target responses differed from those in Phase 1. Overall, the present discrete-trial procedure could produce reliable resurgence as with typical free-operant procedures and has several potential benefits for studying resurgence.

Journal of the Experimental Analysis of Behavior, 2025 · doi:10.1002/jeab.4226