Medidas de Memória em Procedimentos de Equivalência por Pareamentos

Abstract

The equivalence paradigm is one way to understand symbolic behavior from a behavior analytic perspective. The corresponding experimental literature documents the emergence of novel stimulus relations after conditional-discrimination (matching-to sample, MTS) training. Similar emergent performances have been reported after respondent-type training procedures, which do not rely on the operant reinforcement of matching responses. This type of procedure consists of pairing stimuli in the absence of reinforcing consequences, and then testing the emergence of stimulus relation in unreinforced matching-to-sample tests. Thus, both matching-to-sample and respondent type training seem sufficient to induce novel stimulus relations, but the underlying processes remain poorly understood. The main objective of the present study was to evaluate the effects of respondent-type training on the process known as "false memories" (defined as the false recognition of stimulus pairs that were never presented). Thirty-six college students were exposed to stimulus pairings of the A-B, A-C type. Consistent (i.e. similar to training: e.g., B1A1) and inconsistent (i.e. different from training: B1A2) stimulus pairs were then evaluated through recognition tests on a Likert scale. The pairs evaluated in testing could be either direct pairs (i.e., of the BA and CA types: "symmetry") or indirect pairs (i.e., "symmetry combined with transitivity"). The emergence of novel relations was also assessed through traditional matching-to-sample tests. Results showed that respondent-type training can generate correct responses in matching-to-sample, confirming previous findings. Some subjects also showed a tendency toward false recognition of indirect pairs that were never presented in training. These findings suggest that processes of pair recognition and choice responses are partially independent after respondent-type training.