A within-subject comparison of stimulus equivalence training.

Training structures have been defined as the order and arrangement of baseline conditional discriminations within stimulus equivalence training. The three training structures most often used are, linear (trains A:B and B:C discrimination), many-to-one (trains B:A and C:A discriminations) , and one-to-many (trains A:B and A:C discriminations). Each training structure trains a different set of simultaneous and successive discriminations that are then needed in the test for derived relations (symmetry, reflexivity, transitivity, and symmetrical transitivity). The present experiment seeks to extend the research on stimulus equivalence training structures by using a within-subject design and adult human subjects. Three sets of 9 arbitrary stimuli were trained concurrently each with a different training structure. From the beginning, training and testing trials were intermixed. The likelihood of producing stimulus equivalence formation was equal across structures