Peer-Mediated Intervention in Young Children with Autism: Stay, Play, Talk

Background:

The earlier intervention begins for children with autism, the more likely the intervention will guide brain and behavior development back toward a typical pathway (Dawson, 2008), and the more likely children will be included in regular education placements (Harris & Handleman, 2000). However, the efficacy of social communication interventions remains unclear (Rogers & Wallace, 2011). Recent evidence suggests that peer-mediated interventions (PMI) improve social communication skills of preschool and school-age children with autism (Chan et al., 2009; Goldstein et al., 1997, 2007). However, PMIs have not been implemented with children identified with autism in early childhood.

Objectives:

The purpose of this study was to examine the efficacy of the Stay, Play, TalkPMI (Goldstein et al., 1997) on the social communication of children with autism under age 4.

Methods:

Four intervention dyads were created; each included a child with autism (target child) and one typically developing preschooler (peer buddy). Pairs were matched on gender, socioeconomic level, and interests.  Adapted from English et al. (1997), the interventionist taught the peer buddy to Stay, Play, and Talk across three 15-minute pre-intervention training sessions. Peer buddies implemented the Stay, Play, Talk strategies with their matched target child during two weekly 20-minute, sessions across 16 weeks. Each target child’s social communication acts were coded then analyzed across three types: uncoordinated gestures, coordinated gestures, and words with or without gesture. A total communication score (TCS) was created by combining initiation and response data for each participant. The investigators also coded the frequency and type of adult prompting required for the dyads to maintain engagement. Finally, the frequency of social communication acts per child-initiated topic was analyzed to determine which materials or activities were most likely to encourage balanced social communication turns among the target child and his peer buddy.  The Early Communication Indicator (Greenwood, et al., 2006) was used to measure social language use during a 5-min. individual play sample, pre- and post-the intervention phase).

Results:

A multiple baseline design across participants was implemented. Preliminary results are based on the TCS and the frequency of initiations and responses across 6-10 baseline sessions and 8 intervention sessions. Thus far, each child demonstrated stable baselines and a striking increase in TCS at the onset of intervention. Frequency of initiations and responses also increased above baseline data following the onset of intervention. Initiation and response data mirrored the pattern observed for TCS. However, the target children responded to the social communication of their peers more often than initiating it. This pattern fits the social communication profile for children with autism. Type of communication and frequency of adult prompting will also be presented. Video samples highlight the remarkable increase in social communication, as well as the relationships that developed among the peers.

Conclusions:

When used to supplement comprehensive early interventions, peer-mediated interventions show promise in the improvement of social communication in early autism for these children. Results of this study offer valuable clinical implications for the integration of peer-mediated intervention into daycares, homes, and other natural environments.