Synthesized alternative reinforcement and resurgence
Bundled reinforcers in DRA crush problem behavior quickly, but they also fuel a bigger comeback when you remove them.
01Research in Context
What this study did
Smith et al. (2024) compared two ways to run DRA. One group got a single reinforcer for the new behavior. The other group got a bundle of reinforcers at the same time.
The team watched how fast problem behavior stopped and how much it came back when they later removed the rewards.
What they found
The bundle package stopped problem behavior faster and almost wiped it out.
When the team later took the rewards away, the bundle group showed a bigger bounce back than the single-reinforcer group.
How this fits with other research
Johnston et al. (2017) already showed that richer rewards knock behavior down faster but rebound harder. The 2024 study says the same rule holds when you make rewards richer by stacking them instead of just giving more of one.
Greer et al. (2024) worked with real destructive behavior and found that big early cuts in DRA cause the biggest return of problem behavior. Their clinical data line up with the lab finding here: dropping a whole bundle at once is like a big cut, so resurgence spikes.
Smith et al. (2026) took the next step. They showed that removing just one reinforcer from the bundle still sparks a smaller bounce. Together the three papers tell a clear story: the more you pack into DRA, the faster suppression you get, but any removal—partial or total—can bring the old behavior back.
Why it matters
If you use a token, edible, and praise all at once to kill problem behavior, plan for a stronger extinction burst when you later fade. Thin gradually and track resurgence probes. Consider keeping one strong reinforcer in the mix longer instead of dropping the whole bundle at once.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →If your client gets tokens, edibles, and praise together, try removing just one type this week and watch for any spike in the old behavior.
02At a glance
03Original abstract
In treatments based on differential reinforcement of alternative behavior, applied researchers and clinicians often provide multiple, qualitatively different reinforcers (i.e., synthesized reinforcement) rather than a single reinforcer (i.e., isolated reinforcement) contingent on alternative behavior. Some research shows that providing synthesized reinforcement for alternative responses within such treatments produces more rapid and complete suppression of target behavior; however, there is limited research evaluating the durability of these effects during treatment disruptions. Conceptual explanations of resurgence (e.g., resurgence as choice, context theory) suggest that treatments that include synthesized alternative reinforcement may lead to more resurgence of target behavior when alternative reinforcement is disrupted relative to treatments using isolated reinforcement. We evaluated this hypothesis within a three-phase resurgence evaluation. We exposed rats to isolated or synthesized reinforcement for alternative responding in the second phase, and we exposed rats to extinction in the third phase. Synthesized alternative reinforcement produced more rapid and complete suppression of target behavior than did isolated reinforcement in the second phase; however, exposure to extinction following synthesized reinforcement produced more resurgence. We discuss these results in terms of their implications for applied research and their support for current conceptual explanations for resurgence.
Journal of the Experimental Analysis of Behavior, 2024 · doi:10.1002/jeab.4202