Training Structures and the Formation of Equivalence Classes

Abstract

Different training structures have produced different outcomes of equivalence yields when using a simultaneous training protocol. According to the discrimination analysis by R. R. Saunders and Green (1999), the number of simple discriminations required in conditional discrimination differs for the different training structures. Hence, for the MTO (many-to-one) structure, all the simple discriminations are required during training of the conditional discriminations, while they are not in the OTM (one-to-many) or LS (linear series) training structures. The differences in the number of discriminations will be greater when the number of members in the classes increases; thus, there should be a more pronounced difference in the outcome on equivalence tests. The purpose of the present experiment was to train 3 potential classes of 3 members each and 3 potential classes of 6 members each with MTO, OTM, and LS training structures. Thirty adult participants were randomly assigned to 6 groups, that is, 3 or 6 members with MTO, OTM, or LS. The results showed that there were small differences in the outcome following MTO or OTM. However, the equivalence outcome was lowest following the LS training structure. The data from the follow-up tests showed that none of the participants in the LS groups responded in accordance with stimulus equivalence, while 6 of 10 did so in the OTM groups and 3 of 8 did in the MTO groups. Moreover, the reaction time data showed that there was an increase from directly trained trials to the symmetry and equivalence trials, with a more pronounced increase for equivalence trials.