Effects of high, low, and thinning rates of alternative reinforcement on response elimination and resurgence.
Dense alternative reinforcement crushes problem behavior fastest yet rebounds hardest—thinning schedules don’t beat staying lean.
01Research in Context
What this study did
The team compared three ways to deliver alternative reinforcement during extinction.
One group got rich, high-rate rewards. Another got lean, low-rate rewards. The third started rich then thinned to lean.
They watched which plan best stopped the old response and whether it came back when rewards ended.
What they found
High-rate reinforcement stopped the target response fastest. When it ended, the response surged back.
Low-rate and thinning plans weakened the response less, but the comeback was smaller.
Thinning from high to low gave no bonus over staying low the whole time.
How this fits with other research
Johnston et al. (2017) ran a similar rich-versus-lean test with humans and saw the same trade-off.
Shahan et al. (2020) mapped the size of the drop: bigger cuts in reinforcement make resurgence explode.
Irwin Helvey et al. (2023) looked at kids after FCT and found no rate effect—an apparent contradiction. The lab shows rate matters; the clinic, with added communication skills, may soften the impact.
Why it matters
Pick your poison. Want fast suppression? Use dense DRA, but plan for a later spike and guard against it. Prefer a gentle fade? Stay lean from the start and accept slower gains. Either way, thinning schedules give no magic middle road.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →If you run rich DRA, add an extinction burst plan before you fade reinforcement.
02At a glance
03Original abstract
A common treatment for operant problem behavior is alternative reinforcement. When alternative reinforcement is removed or reduced, however, resurgence of the target behavior can occur. Shahan and Sweeney (2011) developed a quantitative model of resurgence based on behavioral momentum theory that suggests higher rates of alternative reinforcement result in faster response elimination and greater resurgence when removed, whereas lower rates of alternative reinforcement result in slower response elimination but are followed by less resurgence. Thus, the present study was designed to examine whether faster target response elimination and less resurgence could be achieved by beginning with a high rate of alternative reinforcement and gradually thinning it such that a low rate is ultimately removed during a simulated treatment lapse. Results showed that high rates of alternative reinforcement were more effective than low or thinning rates at target response suppression but resulted in resurgence when discontinued. Low and thinning rates, on the other hand, were less effective at response suppression but target responding did not increase when alternative reinforcement was discontinued. The quantitative model cannot currently account for the finding that lower-rate alternative reinforcement may not effectively disrupt behavior relative to an extinction only control. Relative advantages of high, low, thinning, or no alternative reinforcement are discussed with respect to suppression of target response rate during treatment, resurgence when alternative reinforcement is removed, and alternative response persistence, while taking into account differences between this animal model and modern applied behavior analytic treatments.
Journal of the experimental analysis of behavior, 2013 · doi:10.1002/jeab.26