Studying stimulus equivalence: Defense of the two-choice procedure

Sidman (1987) and Carrigan and Sidman (1992) have advised against the use of two-choice procedures in studies of emergent matching to sample. They argue that in two-choice, as opposed to multiple-choice, procedures: (1) It is more difficult to make sure that the baseline conditional relations have been established; (2) There is a greater chance that test outcomes are not related to the baseline conditional relations; and (3) The predictions of stimulus equivalence are less clear. In response to the first two arguments, I argue that they refer to technical difficulties that can easily be handled within the two-choice procedure itself. In response to the third argument, I argue that the formulation of Carrigan and Sidman is a new account, that it is inconsistent with the old account of Sidman and Tailby (11982), and that the two-choice procedure goes well with the old account. Further, I argue that there are no strong reasons for adopting the new account, and that the recommendation of using three choices in particular is problematic.