Strengthening Versus Signalling Effects of Reinforcers: A Translational Study

Abstract

There is a growing body of basic research highlighting the shortcomings of a simple response-strengthening account of reinforcers. The effects of reinforcers might primarily be to signal the availability of future reinforcers rather than strengthen behaviours that they follow. The present study extended a basic experimental model, used to isolate strengthening and signalling properties of reinforcers with nonhuman animals, to humans. Nine typically developing children and five children diagnosed with autism spectrum disorder participated in a two-alternative choice game on a tablet computer. In-game reinforcers (treasures) were generally available sooner for the response that had just produced the previous treasure (same-sooner conditions) or for the response that had not produced the previous treasure (other-sooner conditions). Phase 2 had similar conditions but with qualitatively different in-game reinforcers arranged for the two variable interval schedules. Phase 3 included a continuous signal for the most recent treasure location. Reinforcer control was generally weak for both groups of participants across all conditions. However patterns of choice were not consistent with reinforcer-strengthening effects, but suggested weak control by what reinforcers signalled about the likely location of future reinforcers. Potential factors for weak control are discussed.