Signaled reinforcement: Effects of signal reliability on choice between signaled and unsignaled alternatives

• Published in: Behavioural processes 2020-05, Vol.174, p.104088-104088, Article 104088
• Main Authors: Gomes-Ng, Stephanie, Macababbad, Athena C., Bai, John Y.H., Baharrizki, Darren, Elliffe, Douglas, Cowie, Sarah
• Format: Article
• Language: English
• Subjects: Concurrent schedule; Matching; Pigeon; Response competition; Signal reliability; Signaled reinforcement; Stimulus control
• ISSN: 0376-6357; 1872-8308
• DOI: 10.1016/j.beproc.2020.104088

•Stimuli signaling reinforcement reduce responding on one alternative of a concurrent schedule.•We asked how signal reliability affects choice between signaled and unsignaled alternatives.•Signal reliability changed responding on the signaled alternative, but not the unsignaled alternative.•These results are consistent with matching theories, but not with research showing changes in responding on both alternatives.•Response competition and local reinforcer rates may modulate effects of signaled reinforcement in concurrent schedules. When reinforcer availability on one alternative of a concurrent schedule is signaled by a discriminative stimulus, responding on that alternative decreases. We investigated how the correlation between signal presentation and reinforcement (signal reliability) affects choice between signaled and unsignaled alternatives. Six pigeons responded in a concurrent schedule, in which reinforcers on one alternative were signaled by a key-color change. Across conditions, the probability of reinforcement following signal presentation varied (the probability in its absence was the complement). As signal reliability increased, response rates and latencies following signal onset on the signaled alternative decreased, whereas responding on the unsignaled alternative remained unchanged. Because the signal did not alter overall reinforcer rates, these findings are consistent with matching theories and research suggesting that responding on one alternative of a concurrent schedule depends on reinforcer, but not response, rates on other alternatives. However, these findings are inconsistent with others demonstrating concomitant changes in responding on signaled and unsignaled alternatives. We consider whether a response-competition account of concurrent performance can explain these discrepancies, and suggest avenues for future studies to investigate the mechanisms underlying effects of signaled reinforcement in concurrent schedules.