The effects of reinforcement magnitude during token gain and loss on procedural efficacy and preference.

Abstract

Results regarding the efficacy of and preference for token gain and loss within token economies have been inconsistent. The purpose of the current study was to evaluate the influence of reinforcer magnitude on on-task behavior by changing the schedule of reinforcement during token gain and loss procedures. Results showed that when the token schedule was the same, token gain and loss were similarly effective in increasing on-task behavior for most participants; however, three participants preferred the gain procedure and one preferred the loss procedure. When the token schedule was thinned for one procedure and not the other (e.g., five tokens were available in gain and 10 were available in loss), token gain and loss were similarly effective, two participants had consistent preference for gain, and two participants' preference changed when the schedule of reinforcement changed.