Examining the Use of Turn Taking In Pivotal Response Training: Adaptations for Classroom Environments

Background: Pivotal Response Training (PRT) utilizes turn taking as one tool to target communication, play, and social goals for children with autism. The therapist takes a turn in PRT for two reasons: (1) to model appropriate language, play and social skills for the child (modeling) and (2) to provide an opportunity for the child to produce a behavior (independently or with a prompt). When the child engages in the behavior, the adult can then provide reinforcement by returning the materials to the child (contingency). Unlike other components of evidence-based interaction, the use of turn taking with children with autism has not been individually empirically validated in the scientific literature. Teachers using PRT in applied settings have consistent difficulty using turn taking, and often exclude this point in their implementation with students.

Objectives: The function and necessary components of turn taking in PRT will be evaluated and discussed to inform the adaptation of PRT for classroom environments.

Methods: Participants included six children with ASD, ages 3 to 5 years. Children received two, 90-minute sessions of PRT per week for a period of three weeks with a trained therapist. Therapists systematically varied the turn taking component of PRT among four conditions: modeling only, contingency only, modeling and contingency independently, and modeling and contingency together.  All other components of PRT were held constant. Children’s language and play behavior were behaviorally coded for frequency, complexity, function, and type. Results were compared among conditions.

Results: Results indicate that both modeling and contingency components of turn taking are necessary to optimally promote children’s use of more complex language and play skills. For example, preliminary data indicate subjects engaged in approximately 40% more discrete play actions during conditions involving both modeling and contingency than either modeling or contingency alone. Similarly, the total duration of functional play was 30% longer for the combined conditions.  Though both modeling and contingency are necessary, there appears to be limited difference between the simultaneous use of these components and the independent use of both strategies throughout the intervention session.  Additionally, the initial skill level of the child may differentially effect which components are most successful in promoting skill use.

Conclusions: Findings indicate that PRT protocol may be simplified for classroom implementation by separating the turn taking components of modeling and contingency. Future research will test the effect of this PRT adaptation in applied settings, whether separation of the components affects child engagement, and whether teacher’s implementation of PRT is improved by this alteration