Effects of collateral peer supportive behaviors within the classwide peer tutoring program.

A classwide peer tutoring procedure was implemented in an urban elementary school classroom to improve students' spelling performance. Three students combined untrained or collateral tutoring behaviors with the core behaviors initially taught. To explore the function of these natural and spontaneous behaviors, a multielement single-subject experiment with replications was conducted. Results indicated that the additional tutoring behaviors increased (a) the academic response frequencies of 3 tutees and (b) the weekly spelling achievement of 1 target tutee. The remaining class members were successfully taught and continued to use these behaviors over the final 3 weeks of the school year. These findings are discussed with regard to academic instruction, natural communities of peer reinforcement, and the social validation of intervention procedures.