Sorting: An Alternative Measure of Class Formation?

Abstract

In initial sorting tests, 16 participants did not assign stimuli to experimenter-defined classes. Then, the baseline relations for 5-member equivalence classes were trained using matching-to-sample (MTS) trials. Follow-up MTS tests assessed class formation. Regardless of outcome, another sorting test assessed delayed class formation if classes had not formed or class-maintenance if classes emerged during the MTS test. Classes were not formed by 11 participants, emerged on a long-delayed basis in the sorting test for 2, emerged on a delayed basis in the first MTS test for 2 others, and emerged immediately in the MTS test for 3 others. The latter three participants then attempted to form new equivalence classes. After baseline training, emergence was assessed with a sorting test administered immediately thereafter, and was followed serially with an MTS and final sorting test. Responding in the first sorting tests demonstrated the immediate emergence of the stimulus classes for these participants. The MTS test results implied that the classes that emerged in the sorting test were actually equivalence classes. For the two participants who showed delayed class formation, class integrity was maintained during the follow-up sorting and MTS tests. Two other participants showed class formation in the last and final sorting test. The remaining nine did not show class formation. Because the sorting tests were completed 90 % faster than MTS tests, they provided a quick and reliable alternative to MTS tests for the tracking of class formation.