Extinction contexts fail to transfer control: Implications for conditioned inhibition and occasion-setting accounts of renewal.

Abstract

The renewal effect is often explained as a side effect of the extinction context acting as a negative occasion setter. Four experiments tested whether extinction contexts show the selective-transfer property of occasion setters. Experiments 1–3 used a predictive judgment task where participants rated the probability of certain foods (cues) producing gastric malaise (outcomes) in different restaurants (contexts). Experiment 4 used a behavioral suppression task where sensor lights (cues) served as signals to suppress firing responses in certain galaxies (contexts). All 4 (Experiments 1–4) addressed whether a potentially negative occasion-setting context transferred its modulatory power to an extinguished (presumably occasion set) target in the test phase of an ABC renewal design. Experiments 2–4 further assessed the possibility that the extinction context acts as a conditioned inhibitor by testing a simple excitor on a context where extinction occurred. Neither selective (occasion-setting) nor nonselective transfer (conditioned inhibition) was demonstrated. Implications for theories of renewal and occasion-setting are discussed.

Concurrent evidence of extinction making acquisition context specific and ABA and ABC renewal effects in human predictive learning.

“…Specifically, occasion setters typically only modulate responding to their own target cues and other cues trained as targets of other occasion setters (Holland, 1992). One recently published study has indeed reported a failure in the ability of extinction contexts to modulate responding to similarly-trained cues trained in other contexts (Balea et al, 2020). This failure may highlight a fundamental difference between discrimination learning using discrete cues as occasion setters and the modulatory roles our contexts play.…”

Increasing previous but not concurrent extinction attenuates the “extinction makes acquisition context specific” effect in human predictive learning.

“…However, the occasion setting literature with discrete cues shows that for occasion setting to develop a target stimulus needs to be treated different when it is presented in the presence and in the absence of the putative occasion setter (Holland, 1991; Holland & Reeve, 1991). As that condition is not fulfilled in our experiments (P is never presented alone), to assume that contexts play the role of modulators in the EMACS effect would require us to accept that the modulatory role played by the contexts is different from the modulatory role played by discrete cues, as recent studies begin to suggest (see Balea et al, 2020). Alternatively, it is also possible that contextual control on extinction and EMACS effects would be exerted through different underlying mechanisms, even when they are found within the same situation, as reported by Ogallar et al (2021).…”

Effect of the differential discount value on the purchase prediction: an approach from associative learning

“…En la parte superior de la escala, y de manera equidistante, aparecían cuatro etiquetas: nada, poco, bastante, y mucho, escritas en color negro. La escala correspondió a la adaptación de la utilizada por Aristizabal et al (2017); Balea et al (2020); León et al(2011) y Ogallar et al (2019. Una vez el participante respondía, aparecía el resultado (descuento) con una duración de 3000 ms.…”