Eliminating biased selections during concurrent‐chains schedules: A discovery research example

When an individual has limited language, concurrent‐chains schedules offer an objective means of assessing preference among behavior‐change procedures so that recipients can be incorporated into the treatment selection process. We initially attempted to use a concurrent‐chains schedule to assess children's preference for different types of token systems. However, upon observing children engaging in patterned selections, we evaluated procedures for eliminating biased selections. This was done using simple and salient terminal‐link arrangements with seven children diagnosed with autism spectrum disorder. For five of seven children, manipulating the number of successive forced‐choice trials to the reinforcement link was sufficient to produce discriminated responding. Removing the academic task in the terminal link was necessary for two children. For all participants, discriminated responding maintained when the initial exposure arrangement was reinstated. Results are discussed regarding the potential effects of a history of errorless teaching biasing selections during concurrent‐chains schedules.