Stimulus equivalence class formation and stimulus sorting

Abstract

Immediate emergence of equivalence classes has been documented in sorting tests in two experiments. Discordant findings — showing that classes documented by sorting tests immediately after the training of baseline relations cannot always be defined as equivalence classes — have also been identified. To investigate these types of discordant findings further, we reviewed 25 articles. The reviewed articles were selected from searches in PsycINFO and by examining the reference lists in the already identified articles. The inclusion criterion was whether sorting tests were used in relation to MTS-based training or a similar procedure to establish conditional discrimination. Studies that met the criterion were quantified and evaluated along a number of dimensions related to parameters and variables in procedures. We employed Chinese characters as compound stimuli to investigate the variables influencing attending behavior in matching-to-sample (MTS) and sorting tests. The Chinese characters were used with and without the radical. Twenty participants were assigned to two sequences of experimental conditions, participants in both sequences were exposed to pre-sorting of stimuli, first without, then with the radical, and subsequently exposed to a hybrid MTS (H-MTS) training in which the compound stimuli included identical radicals and therefore were only partly arbitrary. After training the baseline relations, the participants in Sequence 1 had the sorting tests, without and with the radical, followed by a H-MTS test, with the radical, and finally they had the post-sorting of the stimuli without and with the radical. In Sequence 2 participants conducted the H-MTS test after training of baseline relations then the sorting tests and finally the H-MTS test with the radical. The immediate emergence of all stimulus classes was seen in seven of the ten participants in both sequences, whereas for the compound stimuli without radicals, it was only seen in two of the ten participants in Sequence 1 and one of the ten in Sequence 2.