Explorando a dinâmica da formação e reorganização de classes de equivalência por meio da velocidade de respostas de escolha

Abstract

Studies that investigated equivalence class reorganization usually adopted participants’ response accuracy as dependent variable. The choice responses speed is sensitive measure to effects not detected by the accuracy and has been used to assess stimulus relatedness degree. We aimed to explore responses’ speed as a measure of the relational strength between stimuli of reorganized classes. In Experiment 1, undergraduates were divided into two groups after trainings and tests of equivalence class formation. The Reversal Group was submitted to a reversal training targeting some relations previously learned, while the Control Group performed a training involving new stimuli. Equivalence relations were then tested again. The Control Group speed increased from the first to the second tests; no differences between tests were observed for the Reversal Group. Experiment 2 investigated whether reorganization test speed is influenced by the amount of training exposure. Two groups underwent trainings and tests of class formation, followed by trainings and tests of class reorganization. An overtraining was performed by the StF Group before the formation tests and by the StR Group before the reorganization tests. The equivalence relations maintenance was tested after 14 or 30 days. The results showed that the StF Group speed was stable throughout the tests while StR Group speed increased from the formation to the reorganization tests and decreased in the maintenance. This indicates that overtraining increased the relational strength of reorganized classes, but impaired their maintenance. Experiment 3 manipulated the inclusion of meaningful stimuli in the classes. The group that related emotion expressions to abstracts stimuli (FACE Group) was more likely to respond correctly in training trials for class formation than the group that only used abstract pictures (ABS Group). When relations were reversed, the FACE Group was less likely to respond correctly during training. Speed showed that the FACE Group responded faster in the formation tests, while the ABS Group responded faster in the reorganization tests, although accuracy was the same for both groups in these steps. Therefore, the use of meaningful stimuli increased the probability of class formation, but hindered class reorganization. Experiment 4 used a pairing procedure to establish meaningful and meaningless S-S relations. A multiple-choice task assessed relational strength between stimuli. The multidimensional scaling technique was used to produce graphical representations of how equivalence classes are arranged in space based on responses speed. The use of meaningful stimuli was investigated, but unlike Experiment 3, the representations did not reveal differences in class formation and indicated that the stimuli were more strongly related after the reorganization for the group that used meaningful stimuli. These findings were interpreted in the light of Relational Density Theory (RDT). In short, the manipulated variables impacted class density and resistance as predicted by RDT, but results such as those of Experiment 4 still require a more elaborate theoretical interpretation.