This study evaluated if relations between components of complex sample stimuli would have controlled conditional responding in identity and arbitrary simultaneous matching-to-sample tasks. In Phase 1, 3 children with educational special needs were trained to select stimulus B1 in the presence of stimulus A1, and select B2 in the presence of A2 (AB conditional relations); then symmetrical BA relations BA were tested. Afterwards they were exposed to differential observing response procedure that prompted children to make simultaneous identity matching responses with complex sample and comparison stimuli (AB-AB relations). In the sequence, the ABX conditional relations were trained. One stimulus in set A and another in set B appeared together as a sample, and two novel stimuli were the comparisons. Selection of X1 was reinforced if the two stimuli in the AB complex sample had been related in the previous training, and selection of X2 was reinforced if the components of AB sample had not been conditionally related. In Phase 2, PQ and QP conditional relations had been trained and tested, respectively. The aim of PQX tests was evaluated if selection of X1 and X2 would have been controlled by conditional or non-conditional relations between P and Q stimuli as complex sample. All children who learned AB conditional relations, showed BA symmetrical emergency, and obtained high accuracy level at differential observing response, namely they demonstrated AB-AB identity matching-to-sample. Differently neither of them met learning criterion on ABX training. Then the experiment was stopped in Phase 1. In addition to literature data, these results demonstrated functional independency between discriminative skills required by the two teaching contingencies of conditional relations with complex stimuli.
Conditional relations; complex stimuli; differential observing response; children with educational special needs
