The control response in assessing resurgence: Useful or compromised tool?

Abstract

Resurgence experiments sometimes include an operandum on which a history of reinforcement has not been experimentally established. The purpose of this control operandum is to rule out a generalized increase in responding when the alternative response is extinguished as being the cause of the resurgent target response. A review of the results of experiments conducted with both nonhumans and humans in which a control operandum was included shows that control-operandum responding is more common in the latter and almost nonexistent in the former. Both the presence and absence of responding on the control operandum, however, are subject to multiple interpretations thereby rendering it a compromised tool. Alternatives to using a control operandum to rule out extinction induction as the basis for resurgence include a preresurgence test control procedure and a differential resurgence procedure.

Examining effects of training duration on humans' resurgence and variability using a novel touchscreen procedure

“…When a previously reinforced response is placed on extinction, other forms of responding can emerge in addition to the resurgence of target responding, often termed extinction‐induced variability (e.g., Cox et al, 2019; see Neuringer & Jensen, 2013; see also Lattal and Oliver, 2020, for a discussion on the use of this term). Assessing the variability accompanying extinction contingencies along with resurgence provides potential insights into behavioral processes contributing to behavioral treatments (see Ho et al, 2018; Liggett et al, 2018; A. Galizio et al, 2020, for discussions).…”

“…Extinction‐induced variability potentially confounds any conclusions that increases in target responding in Phase 3 are due specifically to reinforcement history in Phase 1 (i.e., resurgence; but see Lattal and Oliver, 2020). Sweeney and Shahan (2016) suggested that including one or more inactive control responses throughout all three phases (e.g., an additional discrete manipulandum which never produces reinforcers) can help distinguish resurgence from general increases in variability.…”

“…Lattal and Oliver (2020) suggested that comparable increases in target and control responding during Phase‐3 testing do not preclude resurgence. In assessments of resurgence in which target, alternative, and control responses do not differ topographically (e.g., Sweeney & Shahan, 2016; Bolívar et al, 2017; Cox et al, 2019), one possibility is that control responses are instances of response generalization.…”

“…Lattal and Oliver (2020) argued that, given topographically identical responses, control responses during resurgence tests could be instances of response generalization. Our novel procedure allows for a direct test of this assertion by arranging topographically identical target, alternative, and control responses along a continuum.…”

“…With this arrangement, examining differences in response‐generalization gradients during extinction tests as a function of training duration could provide insight into the influence of the length of reinforcement history for a target response on allocation to the target response specifically versus generalization to responses varying along some dimension when reinforcement conditions change (i.e., extinction). Given the findings assessing reinforcement history on resurgence (Winterbauer & Bouton, 2010; Winterbauer et al, 2013), we expected that more extensive training would (1) result in greater levels of more distally reinforced target responses relative to more recently reinforced alternative responses (see Reed & Morgan, 2006; Sánchez‐Carrasco & Nieto, 2005) and (2) produce steeper gradients of response generalization with a higher peak at the target during extinction testing (see Lattal & Oliver, 2020).…”