Overtraining in Formation of Equivalence classes

When establishing conditional discriminations, differences in the number of training trials can be some of the reason for the discrepancy in equivalence yields. The current study was employed to study how different numbers of trials required in training influenced the participants’ responding in accordance with stimulus equivalence. In two experiments, we trained three 3-member classes in a Linear Series training structure (AB and BC trials) in 30 adult participants. The training of conditional discrimination was divided into 5 phases. For participants in Group 1, there was a minimum requirement of 18 training trials in each phase, i.e., each trial type was presented three times. For participants in Group 2, there was a minimum requirement of 36 training trials in each phase, i.e., each trial type was presented six times. For participants in Group 3 there was a minimum requirement of 54 training trials in each phase, i.e., each trial type was presented nine times. A test block including baseline, symmetry, transitivity, and equivalence trials followed the training. The results showed that in Group 1 two of 10 responded in accordance with stimulus equivalence. In Group 2, four of 10 responded in accord with stimulus equivalence. In Group 3, five of 10 responded in accordance with stimulus equivalence. Reaction time data showed that reaction time increase from the last five baseline trials to the first five test trials. The reaction times were shorter for the participants in Group 2 and 3 compared to Group 1. In a second experiment, we controlled for number of reinforced trials. We presented 18, 36, or 54 trials in a block with no programmed consequences. The results showed that it was a minimal effect of number training trials per ce. It seems like the significant variable in overtraining is number reinforced trials and not number of trials by itself.