The measurement of engagement, or the interaction of a person with their environment, is an integral part of assessing the quality of an intervention program for young children diagnosed with autism spectrum disorder. Researchers and practitioners can and do measure engagement in many ways on the individual and group level. The purpose of this methodological study was to compare three commonly used recording systems: individual partial interval, group momentary time sampling, and group partial interval. These recording methods were compared across three classes of engagement: social, instructional, and non-instructional in a clinical setting with children with autism. Results indicate that group measurement systems were not sensitive to individual changes in engagement when child behaviors were variable. The results are discussed in the context of behavior analytic conceptual systems and the relative utility and future research directions for behavior analytic practice and research with young children in group settings.

Dance is an enjoyable activity that children can engage in across the lifespan. Many children with autism have limited leisure activity, such as dance, and also have challenges in terms of overall health related to physical activity. Previous research suggests that there are both immediate and prolonged benefits of exercise. The purpose of the present study was to evaluate the effects of a group dance period on on-task behavior, social attending, affect, stereotypic behavior, and disruptive behavior of three girls diagnosed with autism. The experimenter employed a reversal to evaluate the effects of a "dance party" on a range of behaviors over time. During dance activities, staff and children danced as a group and were observed before and after the dance period. During baseline there was no dance party. While no differences were found across measures, the children did have high levels of favorable affect during the dance party. The results are discussed in the context of previous literature and directions for future studies.

The differential outcomes effect (DOE) refers to an observed increase in rates of acquisition of simple or conditional relations when the contingencies of reinforcement arrange for reinforcers to be uniquely correlated with a particular stimulus or response relative to conditions where the reinforcers are not uniquely correlated with either stimulus or response. This effect has been robustly documented in the literature with nonhuman subjects. This study asked whether the DOE would be observed with children with autism spectrum disorder (ASD) learning audio-visual conditional relations. Two participants learned two sets of 3 audio-visual conditional relations. For one set, the training conditions arranged for each of the three conditional relations to be uniquely correlated with a particular reinforcing stimulus (the DO condition). For the second set, the training conditions arranged for the same reinforcer to be used for all three audio-visual conditional relations (the NDO condition). Early results show that audio-visual conditional relations were acquired faster under the DO condition relative to the NDO outcomes condition (accuracy in DO condition was 30.8% higher on average than in NDO condition). These data suggest that differential outcomes should be more thoroughly investigated with children with diagnoses of ASD.

The individual functional analysis (FA) is the most widely used method to identify variables that influence the occurrence of problem behavior. Researchers often modify the FA after the analysis reveals undifferentiated responding. The interview-informed synthesized contingency analysis (IISCA) is one of the most substantial variations of the FA that builds upon the FA modifications. We evaluated the use of the two different functional analytic methods, the FA and IISCA, and the subsequent function-based treatment to reduce problem behavior for three children with autism spectrum disorder (ASD). The individual FA and the IISCA demonstrated differentiated responding for all participants. The treatment based on the results from the traditional FA was effective for all children. We discuss the implications of these findings.