Abstract
A straightforward approach to establishing that a stimulus functions as a punisher is to superimpose a single-alternative punishment schedule on a concurrent schedule of reinforcement with equal scheduled reinforcement on the two alternatives. If the stimulus acts as a punisher, there should be a shift in preference away from the punished alternative and toward the unpunished alternative. We replicated this experiment with the Procedure for Rapidly Establishing Steady-State Behavior (PRESS-B) using point-loss as the punisher. Twenty-four participants worked on sets of three concurrent random-interval random-interval (conc RI RI) schedules of reinforcement: one with no punishment (“baseline” condition) and two with punishment where an RI schedule of punishment was superimposed on one of the alternatives. Compared to their respective baseline conditions, all punished conditions showed a decreased median proportion of time allocated to the punished alternative. This successful replication supports the use of point-loss punishment in PRESS-B, which means that researchers can use the procedure to further study the effects of punishment on human operant behavior.