Topografias de controle de estímulos e equivalência de estímulos : variações paramétricas do procedimento de máscaras

Abstract

Stimulus equivalence literature has shown robust data, but with some result variability between participants, with some forming classes promptly and others who do not demonstrate the expected class formation. Matching to sample (MTS) is the commonly used procedure to verify these relationships formation. However, a situation involving experimental MTS is rather complex regarding the number of potential relationships, so that the participant can form classes based on different equivalence relations that those manipulated by the experimenter. For example, responses that are counted as instances of the same relationship may include different control topographies, some responses controlled by relations of selection (when control of the participant's response is given "selecting" the correct stimulus in the presence of a particular model) while others are controlled by rejection (when there is a "rejection" of a incorrect stimulus in the presence of a particular model). One possibility to check and monitor the different control relations during the baseline training is to use special procedures to induce and identify it. One of these procedures is the blank procedure in which a black square is gradually introduced over attempts conditional discrimination baseline so as to occlude participant s sight of it. This way, separately topographies of selection and rejection can be checked, as well as differences in the formation of equivalence relations. AB, BC and CD relations trainings were programmed, each set with three different stimuli (eg: A1, A2 and A3; B1, B2 and B3, etc.). Control only for rejection or only by selection in baseline training conditional relations BC were planned in one of the trainings, and results compared with another training in which all relations were taught either by selection or by rejection. Other parametric variations (nodal distance increase and extensive pre-training role in positive equivalence results using the mask) were also investigated to advance understanding of the variables that may be related to the equivalence formation. The results indicate that learning BC relationship with only one control relation is not sufficient for the equivalence classes ready formation. These are not guaranteed even if only selecting topographies are learned. Have the whole training - ensuring that the participant learns the relations either by rejection or by selection - assures equivalence formation in much more training opportunities. The results indicated that the extensive pre-training was useful to install some prerequisites behaviors required in the task pairing. Participants who passed the pre-training before starting the experiment were partially successful in achieving the criteria for baseline conditional relations learning training, but the generalized conditional responding did not occur as readily as expected, although it seems to have occurred faster for participants who had gone through extensive pre-training.