Stimulus Equivalence Using a Respondent Matching-to-Sample Procedure with Verification Trials

Abstract

Despite systematic demonstrations of the effectiveness of the respondent-type procedure in producing relational responding, matching-to-sample continues to be the predominant approach for studying equivalence class formation. The length of exposure to stimulus pairings during respondent-type training, and repeated alternation of blocks of training and testing are factors that seem to undermine the practicality and the perceived effectiveness of the procedure. Further, the presentation of stimulus pairs without the simultaneous presence of nonclass members may be said to lack the complexity of the matching-to-sample stimulus arrangement. The present study evaluates the efficiency of a respondent matching-to-sample arrangement after exposing participants to a reduced number of training trials per baseline relation. Probes were added at the end of each block of training trials to verify the strength of S–S relations while keeping the number of acquisition trials to a minimum. At least half of the participants responded correctly to all probe trials per baseline relation after only one exposure, and all participants performed accurately in tests of derived relations. Procedural components of S–S training such as intermixing probes during training and arranging an environment-elicited orientation response towards the relevant stimuli, are discussed as key elements of efficient respondent equivalence training procedures.