Alternative‐reinforcer magnitude effects on resurgence across successive relapse tests in mice
Large alternative reinforcers work fast but rebound hard—expect more resurgence and plan extra checks.
01Research in Context
What this study did
The team worked with 12 mice in a lab cage. Each mouse first learned to press a lever for food pellets.
Next they taught the mice a new way to earn food—nose poking a hole. Some mice got one small pellet. Others got three big pellets.
After the nose poke was strong, they stopped all food. They then watched if the old lever press came back. They ran this test four times in a row.
What they found
Big pellets stopped the lever press faster. The mice switched to nose poking sooner.
But when all food ended, the lever press bounced back harder in the big-pellet group. Small pellets gave a softer rebound.
The same pattern showed up in every relapse test. Bigger backup rewards mean bigger relapse later.
How this fits with other research
Greer et al. (2024) saw the same rule in kids. When they cut rich rewards fast, destructive behavior came back strong.
Fisher et al. (2019) also found that richer past rewards raise relapse risk. The mouse data now match what we see in children.
Nist et al. (2021) showed that thinning rewards slowly lowers relapse. The new study adds that starting with smaller rewards also helps.
Why it matters
If you use big rewards to stop a problem behavior, plan for a bigger bounce-back. Add extra relapse probes and fade the reward size or rate slowly. This keeps gains steady when treatment gets tough.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Start your next DRA case with smaller or thinner rewards and schedule at least three post-extinction probes.
02At a glance
03Original abstract
Alternative-reinforcement-based treatments are common strategies for reducing maladaptive behavior in humans. When conditions of alternative reinforcement are made worse in some way, however, behavior that was targeted for elimination may resurge. Previous research using rat subjects has demonstrated that high-magnitude (relative to low-magnitude) alternative reinforcement produces faster elimination of target behavior but more resurgence once removed. The purpose of the present experiment was to assess cross-species generality of these effects to mice. During Phase 1, lever pressing produced single food pellet reinforcers. Next, during Phase 2, lever pressing was extinguished and groups of mice experienced either small-magnitude (one pellet), large-magnitude (three pellet), or no alternative reinforcement for nose poking. All food was suspended in Phase 3 to assess resurgence. As an additional goal of this study, changes in resurgence across successive determinations were assessed by cycling between periods during which alternative reinforcement was present or absent. Large-magnitude alternative reinforcers produced faster suppression of target behavior but more resurgence when removed than did small-magnitude alternative reinforcers. Moreover, this effect endured across repeated resurgence tests.
Journal of the Experimental Analysis of Behavior, 2026 · doi:10.1002/jeab.70070