Effects of Between-classes Negative Relations Training on Equivalence Class Formation across Training Structures

Abstract

This study compared three three-choice matching to sample (MTS) procedures on their probability to form equivalence relations for each of the three training structures: one to many (OTM), many to one (MTO), and linear series (LS). All three MTS procedures established the same within-class positive relations, but they differed in the number of between-class negative relations included in their baseline trials, with two in the standard (STD) matching procedure, one in the semi-standard (SEMI) procedure, and none in the altered (ALT) procedure. Positive and negative control baseline relations were assessed by using trials including novel stimuli or a blank comparison stimulus before symmetry and equivalence performance testing. The matching procedures did not differ in the rate of positive control that they yielded, but they did in the rate of negative control. The number of between-class negative relations in the baseline trials was directly related to the probability of equivalence class formation. However, this relation was different across training structures: it was lineal for the OTM structure, did not differ between the STD and SEMI procedures for the MTO structure, and did not differ between the SEMI and ALT procedures for the LS structure. Testing of positive baseline control by means of a blank comparison stimulus as an incorrect choice and testing of the negative baseline control by using novel stimuli as the correct choice proved more sensitive to the matching procedures and most discriminative between participants who formed equivalence relations and those who did not.