Ensino de relações arbritárias e busca de simetria em Cebus apella

Abstract

For some theorists, the lack of control over procedural variables has made difficult documenting equivalence classes in nonhuman subjects or human participants with developmental disabilities. The present research was divided in two experiments. The Experiment I aimed to evaluate the combined use of procedures such as “specific reinforcement”, “variations of S-“, “blocks of trials gradually smaller”, and “omission of previous S+ on the next trial block” in arbitrary relations training and symmetry test. A young male capuchin monkey (Cebus apella), Guga, served. The procedure comprised the following phases: pre-training, simple discrimination training, arbitrary relation training (A1B1 and A2B2), and symmetry test. The results show that Guga finished the arbitrary training in 39 sessions. The results of symmetry test show accurate performance for one relation and chance level accuracy for the other, with accurate performance in the first test trial for both relations. In Experiment II, two new relations (A3B3, A4B4) were added to the baseline (A1B1, A2B2). The results of the symmetry test for the new relations show accurate performance concerning to one relation and 75% of accuracy for the other. The present study encourages the development of procedures to reduce stimulus control incoherence and reinforces the possibility that the difficulty to document equivalence class formation in nonhumans relays on procedural variables.