This study assessed the impact of choice making on the serious problem behaviors of 3 students with severe autism and/or mental retardation. In the context of within-subject reversal designs, the results showed consistently reduced levels of problem behaviors (e.g., aggression) when the students were given opportunities to make choices among instructional tasks and reinforcers. Additional data showed no systematic differences in the rate of correct responding between the two conditions. The results are discussed in relation to the continuing search for effective, nonintrusive

One of the characteristics of autistic children is severe social avoidance behavior. We assessed whether the type of activity (child-preferred vs. activities that were arbitrarily determined by an adult) engaged in during an interaction was correlated with the amount of social avoidance behaviors these children exhibit. Results revealed a negative correlation between appropriate child-preferred activities and social avoidance behavior. Additional analyses revealed that (a) social avoidance behaviors could be manipulated within a reversal design, and would predictably decrease when the children were prompted to initiate appropriate child-preferred activities; and (b) these procedures could be used to teach children to initiate child-preferred activities in community settings, resulting in reductions in social avoidance responses even after the therapist's prompts were completely removed. These data suggest that the manipulation of task variables may influence the severe social unresponsiveness that is characteristic of autistic children.

A major problem encountered with autistic children is their characteristic self-stimulatory behavior, which frequently interferes with on-task responding and other appropriate behaviors. However, the experimental literature suggests that with many populations, increased physical activity might positively influence subsequent responding. Therefore, the purpose of this study was to investigate the use of increased physical activity (in this experiment, jogging) as a possible method of decreasing subsequent self-stimulatory behaviors as well as increasing subsequent appropriate responding. Seven autistic children with exceptionally high levels of self-stimulatory behavior participated in the investigation. Self-stimulatory and appropriate behaviors were measured both before and after jogging in a repeated-reversal design. The results demonstrated the following: (1) Brief jogging sessions produced decreases in subsequent levels of self-stimulatory behaviors and also produced increases in appropriate play and academic responding; (2) These changes after jogging were evident in three different experimental settings: during academic responding on preschool level tasks' in a clinic; during ball-playing in an outside play area; and in a quiet room, while no other activity was occurring; (3) Supplementary measures obtained in an applied classroom setting showed a similar relationship with both increases in on-task activity and general interest ratings for school tasks following the jogging sessions.