Stimulus Sorting: A Quick and Sensitive Index of Equivalence Class Formation

Abstract

Fifty college students were given standard matching-to-sample-based training and testing to form three 5-member equivalence classes that had linear series training structures (A→B→C→D→E). A matching to sample (MTS) based derived-relations test was used to assess the emergence of the classes. Thereafter, 15 cards, each of which contained one of the 15 stimuli in the classes, were given to the participants who were asked to sort them into piles of “related” cards. There was a very high concordance between measures of equivalence-class formation produced by relation tests and sorting tests. In some cases, sorting tests appeared to be more sensitive to tracking class formation than the MTS-based derived-relations tests. In addition, sorting tests were administered in a fraction of the time needed to administer an MTS-based derived-relations test to document class formation. Finally, sorting test data were very easy to interpret, especially when findings other than experimenter-defined class formation was produced. Therefore, sorting may be a viable alternative for tracking equivalence class formation.