Differential effects of token production and exchange on responding of children with developmental disabilities

•Token economies are commonly used in educational programming despite limited understanding of underlying mechanisms.•Past research suggests token delivery may actually suppress responding among hon-human animals in some contexts.•We compared the responding of three participants under 1) dense token delivery, 2) lean token delivery, and 3) yoked reinforcement conditions.•Given high schedule requirements, token delivery conditions corresponded with lower levels of responding relative to yoked reinforcement.•In the context of high schedule requirements, token schedules may suppress responding by signaling long delays to backup reinforcement. Token economies are commonly used in educational and clinical settings as tools for reinforcing appropriate behavior. However, little applied research has been conducted to investigate the behavioral mechanisms involved. What’s more, research with non-human animals suggests that tokens may serve discriminative functions which may actually suppress responding under high schedule requirements. This study explores the relationship between schedules of token production (i.e., dense, lean, and yoked reinforcement) and response patterns in three children with autism spectrum disorders (ASD). Relative to yoked reinforcement schedules, token production corresponded with higher response rates and shorter pre-ratio pauses (PRPs) under low schedule requirements. Under high schedule requirements, all three participants demonstrated higher response rates and shorter PRPs in the yoked reinforcement condition relative to either token production condition. These findings are consistent with research involving non-human animals suggesting that, under high schedule requirements, tokens may signal lengthy delays to reinforcement and potentially even suppress responding. Thus, we provide preliminary recommendations for using token reinforcement systems most effectively in applied contexts.