Identity and oddity matching are relations derived from conditional discrimination processes that can be studied through matching-to-sample procedures. Generalized matching happens when the subject’s response is controlled by these relations even when novel stimuli are presented. The aim of this study was to evaluate the moment when the identity and oddity relations are generalized, considering how many pairs of stimuli are needed for demonstrating generalization. Six male Wistar rats were trained on an oddity matching-to-sample procedure, and four were trained on an identity matching-to-sample. The olfactory stimuli were cups filled with scented sand in which the rats should dig to obtain a sugar pellet according to either the identity or oddity contingency. The experiment was divided into three phases, each having the training of one pair of stimuli and, after meeting the training criteria, a generalization test with novel stimuli. For this experiment, generalization is considered if the subject matched consistently according to the trained relation in at least 91,67% of the trials during the last testing session of each phase. The experiment ended when generalization was demonstrated in any of the three phases. The applied training showed to be more effective for oddity than for identity relations since the oddity group learned the trained relations faster and five of the rats were able to demonstrate generalization according to the defined testing criteria. Also, within the oddity group, there were rats that demonstrated generalization with less training than three pairs of stimuli, indicating that training with a large number of stimuli is not necessary to show generalized oddity relations. For the identity group, the number of training sessions was higher and only one rat met the testing criteria. Another problem addressed in this study is the effect of the contingency used during testing sessions over the subject’s performance.
Tests under extinction decreased the choice response rate after a few trials. Tests under an undifferentiated reinforcement contingency solved this problem, however, they don’t eliminate the possibility of discrimination between training and testing sessions, what also could impair the evaluation of generalization. The results indicate the need for further investigations on better ways to evaluate generalization and on the difference in the acquisition of the oddity and identity relations.
Identidade e singularidade são relações obtidas através de processos de discriminação condicional que podem ser estudados utilizando o procedimento de “matching-to-sample” (MTS).
Quando o sujeito responde sob controle dessas relações diante de estímulos novos, tem-se identidade e singularidade generalizadas. O objetivo deste estudo foi avaliar, de forma mais precisa, o momento em que as relações de identidade e singularidade são generalizadas. Para isso, seis ratos Wistar machos foram submetidos a um treino de singularidade e quatro, a um treino de identidade, no qual eles deveriam cavar em copos de areia contendo os estímulos olfativos em procedimento de MTS. O experimento foi dividido em três etapas, cada uma contendo treino de um par de estímulos e teste de generalização com pares não-treinados. No procedimento de singularidade, dois ratos demonstraram generalização após o treino de um par de estímulo, um após o treino de dois pares, e dois após o treino dos três pares; um sujeito não atingiu o critério de generalização. No procedimento de identidade, apenas um dos quatro sujeitos demonstrou generalização. De uma forma geral, o treino aplicado foi mais eficaz para relações de singularidade do que de identidade e foi possível demonstrar generalização utilizando poucos estímulos e poucas sessões de treino.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados