FREE-CHOICE PREFERENCE WHEN ONE ALTERNATIVE IS RARELY OR NEVER CHOSEN
DOI:
https://doi.org/10.18542/rebac.v1i1.761Abstract
In free choice, two or more responses eligible for reinforcement are concurrently available, as when a pigeon’s pecks on either of two keys can produce fixed-interval (FI) reinforcers. In forced choice, only one eligible response is available, as when pecks on one key can produce FI reinforcers but extinction (EXT) is arranged for pecks on a second key. Free choice is typically preferred when pitted against forced choice in terminal links of concurrent-chain or multiple concurrent-chain schedules. When multiple concurrent-chain schedules arrange conditions A and B respectively for left and right terminal links during one initial-link stimulus but their reversal during a second initiallink stimulus, preferences can be determined within sessions as differences between relative initial-link rates. The experimental question was whether free-choice preference is demonstrable even with one free-choice alternative rarely or never chosen. A history of multiple concurrent-chains with equal single-FI terminal links was followed by training,independent of initial links, of FI 20-s (green key), FI 40-s (yellow key), and EXT (red key). Multiple concurrentchain schedules then pitted free-choice terminal links with green (FI 20-s) and yellow (FI 40-s) keys against forcedchoice terminal links with green (FI 20-s) and red (EXT) keys. These terminal links maintained responding almost exclusively on the green key whether the other key was yellow or red, and all reinforcers were produced by green-keyresponding. Even with reinforcers equal and with the yellow alternative rarely or never chosen, the green-yellow erminal link (free choice) was preferred to the green-red (forced choice) terminal link.Key words: free choice, forced choice, preference, concurrent-chain schedules, initial links, terminal links, iscriminated operantsDownloads
Published
2012-03-15
Issue
Section
Research Articles
License
Os autores dos manuscritos aceitos e que não precisarem realizar modificações substanciais devem enviar ao Editor uma cópia digital da verão final acompanhada por carta assinada por todos os autores, indicando que os direitos autorais de reprodução e divulgação do artigo serão da Revista Brasileira de Análise do Comportamento.