Sensory Reinforcement with Autistic Children

Three non-verbal, autistic boys of 9, 12 and 16 years served as subjects in two experiments and a nurse play-therapist acted as the agent. The first experiment compared the effects upon a motor coordination task of each subject's preferred sensory reinforcer with those of the sensory reinforcer preferred by the other two subjects. On-task performances were maintained by prompting and by contingent presentation of each reinforcer in a multiple-baseline design across subjects. All subjects showed higher levels of on-task performance for their preferred sensory activity. In the second experiment a multielement-baseline design compared the effects of the preferred sensory reinforcer with those of a preferred edible reinforcer. All subjects showed higher levels of on-task performance for their preferred sensory activity. Inter-observer reliability remained above 90%. A role for sensory reinforcement in training autistic children is suggested.