Association for Behavior Analysis International

The Association for Behavior Analysis International® (ABAI) is a nonprofit membership organization with the mission to contribute to the well-being of society by developing, enhancing, and supporting the growth and vitality of the science of behavior analysis through research, education, and practice.


49th Annual Convention; Denver, CO; 2023

Event Details

Previous Page


Symposium #401
Basic Evaluations of Resurgence and Spontaneous Recovery
Monday, May 29, 2023
4:00 PM–5:50 PM
Hyatt Regency, Centennial Ballroom D
Area: EAB; Domain: Basic Research
Chair: Sean Smith (SUNY Upstate Medical University)
Discussant: William Sullivan (Golisano Children's Hospital & Center for Special Needs; SUNY Upstate Medical University)
Abstract: Treatment relapse—the recurrence of a previously decreased response—is a complex phenomenon that can be caused by myriad variables. Basic research is paramount for carefully controlling these variables to determine how they interact to influence treatment relapse. In this symposium, we present basic research experiments on resurgence and spontaneous recovery. Each experiment included the implementation of complex schedules of reinforcement, which produced sophisticated histories of reinforcement, prior to evaluating response patterns during extinction. Two presentations feature research analyzing resurgence following complex histories of reinforcement for target responding during baseline. A third presentation includes research evaluating resurgence after a complex history of reinforcement for alternative responding during the second phase of a three-phase resurgence evaluation. A fourth presentation features research analyzing spontaneous recovery after exposure to concurrent VI schedules and varied delay periods. Collectively, these presentations demonstrate how complex histories of reinforcement can be integrated into models of treatment relapse to provide additional insights into the basic principles governing relapse phenomena.
Instruction Level: Intermediate
Keyword(s): relapse, resurgence, spontaneous recovery

On/Off Alternative Reinforcement and Resurgence With Human Participants

SEAN SMITH (SUNY Upstate Medical University), Brian D. Greer (Rutgers Robert Wood Johnson Medical School)

Resurgence is a form of treatment relapse that often occurs when a previously decreased target response recurs due to a decrement in the reinforcement conditions for engaging in an alternative response. Recent research with nonhuman animals demonstrated that cycling between the availability and unavailability of alternative reinforcement within a differential reinforcement of alternative behavior procedure will mitigate resurgence during subsequent exposures to extinction (Shahan et al., 2020). We evaluated whether this cycling procedure would have a similar effect on human behavior by recruiting participants through Amazon’s Mechanical Turk platform to complete a computer-based, human-operant, three-phase resurgence evaluation. Across groups, we varied the differential reinforcement of alternative behavior procedure in the second phase: for certain groups, alternative reinforcement cycled between availability and unavailability (cycling), whereas other groups experienced consistent alternative reinforcement conditions throughout the second phase (constant). Resurgence of target responding occurred to a lesser extent for participants in the cycling groups relative to the constant groups. We discuss these results in terms of their implications for informing clinical practice, evaluating the predictions of Resurgence as Choice in Context with human participants, and recruiting participants on Amazon’s Mechanical Turk platform.

Effects of Testing Delay on Spontaneous Recovery of Choice
JACK VAN ALLSBURG (Utah State University), Timothy A. Shahan (Utah State University)
Abstract: Spontaneous recovery of choice occurs when organisms revert to a response distribution appropriate to a reinforcement distribution from the more distant past, rather than reflecting more recent distribution of reinforcers. By providing a way to examine how organisms weigh and integrate changing relative reinforcement over time, study of the phenomenon may aid in developing improved accounts of other relapse phenomena involving changing reinforcer distribution, (e.g., resurgence). In a series of experiments employing between-groups comparisons, rats responded on two concurrent variable interval schedules. In phase 1, the reinforcer distribution heavily favored option A. In phase 2, reinforcement conditions reversed to favor option B. Rats experienced a delay (during which, no sessions were conducted), and group assignment determined delay length. Following the delay, preference was assessed in test sessions where both levers were presented without reinforcement. Exponentially weighted moving averages (including versions with single and multiple timescales) and versions of the Temporal Weighting Rule were evaluated as potential mechanisms of how reinforcement is integrated over time. None of the models effectively described the effects of delay on spontaneous recovery. An apparent trend toward indifference as delay increases may result from a tendency toward exploration or a decrease in discriminability of past experiences.

Resurgence of Responses Maintained by a Chained Schedule

KENNON ANDY LATTAL (West Virginia University), Firdavs Khaydarov (West Virginia University), Patrick Duffy (West Virginia University), Hiroto Okouchi (Osaka Kyoiku University)

In two experiments, pigeons’ key pecks first were reinforced on a chained variable-interval (VI) 30-s VI 30-s schedule in which initial- and terminal-link responses were reinforced on, respectively, the left and right response keys. This training phase was followed by one during which (a) responding on the left and right keys was extinguished, with both keylights on throughout each session, (b) responding on a center key was reinforced according to a VI 30-s schedule, and (c) in Experiment 2, a change-over delay (COD) was added during both the first and second phases such that pecks to keys incapable of producing reinforcers initiated a 3.5-s changeover delay that prevented active-key responses from being reinforced during the COD. In the final, resurgence test, phase, all three response keys were transilluminated, and each was associated with extinction. Key pecks almost always resurged first on the terminal-link key, often followed later by a small, less systematic, resurgence of initial-link pecks. There were more key pecks during the resurgence test to the terminal-link than to the initial-link key. The results extend hierarchical resurgence effects to chained schedules.

A Resurgence Test of Concurrent Switching Patterns
STEPHANIE P. DA SILVA (Columbus State University)
Abstract: Resurgence of changeover patterns was studied in three pigeons first exposed to multiple concurrent VI 60 VI 60 schedules. In one component (marked by green or red side keys), a 3-s changeover delay operated that "discouraged" rapid switching among the side keys; in the other component, a 3-s switch-over contingency operated that "encouraged" frequent switching among the side keys. This baseline training remained in effect until responding and changeover rates in each component were stable. In the second/elimination condition, the VI 60 schedules were replaced by extinction on the side keys and a center white key was introduced on which a VI 15-s schedule operated. Once pecking on the center key was established and side-key responding was eliminated, extinction was implemented on the center key and resurgence of side-key responding was measured. Of particular interest, was whether the pattern of resurgence in extinction would reflect the baseline patterns as marked by switching among the side keys. Resurgence occurred for two of three pigeons, with one pigeon demonstrating unequivocal differences in switching across components that reflected switching during baseline.



Back to Top
Modifed by Eddie Soh