Reorganization of equivalence classes: effects of preliminary training and meaningful stimuli

In Condition 1, adults learned the baseline relations for the three equivalence classes A1-B1-C1-D1-E1, A2-B2-C2-D2-E2, and A3-B3-C3-D3-E3. Classes contained abstract shapes in the ABS and four preliminary training groups. Each class in the PIC group contained one picture and four abstract shapes. Before class formation for four other groups, preliminary training involved establishing identity (CC) or arbitrary (CX) relations either with or without a delay. Without preliminary training, classes formed with low and high likelihoods in the ABS and PIC groups, respectively. Preliminary training with no delay produced modest increases in class formation, while preliminary training with delay produced large increases in class formation. Condition 2 replicated Condition 1 but with training of reassigned BC and CD relations that linked C from one class to B and D from another class: B1-C2, B2-C3, B3-C1, C2-D1, C3-D2, and C1-D3. Subsequent tests assessed the emergence of the reorganized classes A1-B1-C2-D1-E1, A2-B2-C3-D2-E2, and A3-B3-C1-D3-E3. All preliminary training procedures increased likelihood of forming the reorganized classes to the level seen in the PIC group. Greater gains were produced by preliminary training with no delays than with delays. Test performances also showed how preliminary training influenced baseline acquisition speed and participant-defined relations.