The aim of this investigation was to assess the formation of conditional discriminations by means of a procedure involving complex stimulus training with delay. In phase I five university students received a pretraining TX trial followed by AE, BF and DC acquisition training involving matching-to-sample and simple comparison. In phase II a pretraining N-O/P, the training AB-E/F, and AD-C/F, with consistency training. These training trials were interspersed with EF/FE, CF/FC, and BD/DB transitivity and equivalence tests following complex matching-to-sample and simple comparison. In phase III AB-E/F and AD-CF trials occurred jointly with AB-E, AB-F and AD-C sequences, followed by transitivity and equivalence tests. All of the respondents formed conditional relations involving simple and complex matching-to-sample, albeit with variability across the trials. Variation was also present in the formation of emergent relations. These results suggest that, in spite of the variability, the procedures produced conditional discrimination leading to the formation of equivalent stimulus classes. In future research the identification of the most efficient forms of stimulus control is deemed necessary.
consistent matching-to-sample training; complex stimuli; delayed matching-to-sample
