Date of Award

Spring 2006

Degree Type

Thesis - Restricted

Degree Name

Master of Arts (MA)



First Advisor

Moyle, Maura

Second Advisor

Erdman, Kathleen

Third Advisor

Long, Steven


The communicative deficits exhibited by children with autism are of great concern to parents and educators (e.g., Wetherby, Prizant, & Schuler, 2000). The most commonly used intervention techniques for increasing communicative behavior in children with autism can be categorized as either adults as agents of intervention or peermediated intervention. Debate in the literature exists concerning which intervention approach is most beneficial. The purpose of this study is to compare an adult as intervention agent with peer-mediated intervention for a child with autism. It is hypothesized that a peer-mediated approach will be more effective in terms of increasing the frequency and proportion of appropriate communicative behaviors (i.e., spontaneous verbalizations), increasing the rate of engagement in communicative interactions (i.e., circles of communication), and decreasing the frequency of negative behaviors in general. Consistent with the original prediction, peer-mediated intervention resulted in an increased proportion of spontaneous verbalizations by the child with autism as compared to the adult as an agent of intervention. Also as predicted, peer-mediated intervention resulted in significantly fewer negative behaviors than during intervention utilizing an adult as an agent of intervention. Contrary to the original hypotheses, results indicated that the novice produced more spontaneous verbalizations during the phases including the adult as an agent of intervention rather than the peer-mediated phases. Similarly, significantly more circles of communication occurred during the adult as an agent of intervention, however; the number of circles of communication between the novice and expert significantly increased from the first peer-mediated phase to the second. Implications of these findings are discussed.