Interventions that Increase School-Age Children’s Peer Interactions
Another group of studies has focused on increasing peer interactions among school-age children. Adult instruction in social skills, peer mediation, and peer tutoring are the main methods used in this age group.
One creative adult instruction approach to mixed groups of children used children’s obsessive interests as the basis for invented games. Baker, Koegel, and Koegel (1998) reported a study of three high-functioning elementary-age students. An adult created group games from each of these children’s special interests and then taught the game to the child and some typical peers during a playground period at school. Results included dramatic increases in peer interaction that were maintained through the intervention and follow-up period, generalization to other activities, and increases in positive affect in both target children and peers during interactions.
A particularly well-publicized intervention involving adult instruction involves Gray’s (Gray & Garand, 1993) social story technique, in which written stories are created to teach social rules and the reasons for them in a supportive, noncritical, and understanding way. While the technique is widely discussed and recommended, little empirical work has been done with it. One published study of a single subject has examined the use of social stories to increase social interactions of a child with autism and peers. Norris and Dattilo (1999) used a multiple baseline approach to monitor changes in three social behaviors (increasing appropriate social behavior, decreasing inappropriate behavior, and decreasing isolation) of an 8-year-old girl with autism. Three different stories were written per manual directions to increase social interactions in school at lunchtime. One of the stories was read daily before lunch for 13 days following baseline. There appeared to be a treatment-related change in only one of the three behaviors—decreasing inappropriate behavior—and the response was quite variable. Thus, this technique did not appear to be as effective as others reported above, although this is only a first trial of this technique and more investigation is clearly needed.
In an early paper, Strain et al. (1979) used peertraining techniques in play organizers with four lower functioning elementary-age students. A typical peer was taught to elicit, prompt, and reinforce social behavior between each of two pairs of children with autism. Social behaviors increased dramatically during the treatment phase and fell to baseline when the specific interventions were withdrawn and only general social interactions were used. Thus, there was no evidence of generalization, although later studies of this approach have demonstrated the importance of using typical peers in natural contexts, with careful use and fading of reinforcers for the peers in order to achieve maintenance and generalization.
Shafer, Egel, and Neef (1984) used a direct model and prompt training procedure to train peers who had mild disabilities to elicit increased levels of interactive toy play for four children with autism. Peers rapidly learned the interventions and children with autism demonstrated increased responses and initiations that were maintained over time. Furthermore, when the trained peer and child with autism were combined with two other untrained peers for play sessions, the untrained peers increased their interactions with the children with autism as well, demonstrating generalization across peers. Improved social interactions of both peers and children with autism were maintained over time and generalized across settings. However, both groups tended during generalization episodes to deliver the interventions only to specific children with whom they had been trained.
Lord and colleagues have contributed a program of research using group designs to examine the effects of typical peers on the social engagement of children with autism. Lord and Hopkins (1986) examined several variables related to peer interactions for six schoolage children with autism. They demonstrated that sameage peers elicited more social behavior than did younger peers and that daily exposure in peer play sessions significantly increased a number of social behaviors for the children with autism, including proximity, appropriate play, time spent looking at a peer, and time engaged socially, effects that generalized to new peers. Social initiations were not particularly affected by the intervention, however. These results were replicated in a study of two higher functioning children (Hopkins & Lord, 1981, reported in Lord, 1984).
Lord then conducted a study in which peer training techniques similar to those used by Strain and associates (Strain et al., 1979) were applied. The trained peers facilitated the social engagement of their peers with autism so rapidly that in their first play session, the children with autism were as engaged as they had been in the tenth session with untrained peers, although again, social initiation was much less affected by the interventions. Generalization to unfamiliar trained peers was excellent, though generalization to untrained peers was not as dramatic. A study by Dewey, Lord, and Magill (1988) examined the effects of differing types of play materials on peer play in dyads including a typical child and a child with autism. This study examined both school-age children and adolescents in a day camp setting that lasted 2 weeks. The authors demonstrated that rule-governed games facilitated the most complex social interactions, were the most fun, and kept the children most involved with the interaction.
Construction materials were the next most effective in facilitating more complex interactions, compared to dramatic play and functional play. Finally, two studies of peer behavior in a 2-week day camp experience were examined (Lord & Magill-Evans, 1995). As was found in the first two studies, daily exposure to typical peers who were motivated to interact, but were untrained, resulted in increased social engagement and responsiveness, as well as a decrease in purposeless activity, and an increase in constructive play, for eight children with autism. The second study examined characteristics of social initiations of the children with autism compared to peers, but did not include change data and thus is not reviewed here. In summary, the studies by Lord and colleagues have demonstrated that ongoing, dyadic interactions with a motivated same-age peer results in improvements in social engagement and purposeful play, as well as a reduction in atypical behaviors, for children with autism. The authors hypothesize that the amount of structure, as determined by the characteristics of play materials and by the skills of the peer-play partner, facilitates social engagement by children with autism. Trained peers can create the effects much more efficiently and rapidly than untrained peers.
Peers As Tutors
Kamps and her colleagues have contributed several important studies in this area. Kamps, Barbetta, Leonard, and Delquadri (1994) examined the effects of peer tutoring versus teacher-led instruction on reading performance for three high functioning children with autism. Using a multiple baseline with reversal design, students were paired with classmates for 30 minutes of a specific approach to peer tutoring in which they had all been trained. This approach involved shared responsibility for skill review, correction, and praise in a reading curriculum. Instruction in both conditions was followed by 15 to 20 minutes of free time in classroom centers. Teacher-led teaching (baseline) and peer tutoring conditions were alternated twice, with improved academic performance as well as increased social interaction in free time following peer-tutoring conditions.
Dugan et al. (1995) extended the above findings by examining the effects of cooperative learning groups for two high-functioning fourth graders and 16 typical peers using an ABAB design. Groups consisted of 40-minute social studies learning groups consisting of four children. Groups were held for 3 weeks and alternated with classroom teacher instruction. All members of the groups were trained in both teaching and social interventions. Results demonstrated gains in academic engagement, academic performance, and social interaction during the learning situation compared to the teacher-directed condition for all children involved, including typically developing children. The approach has been replicated but needs long-term follow-up to determine generalizability and long-term gains.
Social Skills Groups
Kamps, Leonard, Vernon, Dugan, and Delquadri (1992) used social skills instruction in play groups consisting of both 11 typical peers and 3 high-functioning first graders with autism. Four group daily play sessions were held, each group consisting of one target student and three peers and some type of planned activity. Social skills training was conducted for the first 10 minutes, followed by 10 more minutes of play in a center activity. After the entire curriculum was taught, feedback was provided for all students during play groups. Follow-up consisted of several social skills groups held at the end of the school year. Significant increases in social skills, length of interactions, and consistency of responding were found for all target students with good maintenance across time.
Pivotal Response Training
Pierce and Schreibman (1997) taught eight typical peers to use PRT techniques to involve two children with autism in improved social interaction during toy play. Multiple baselines demonstrated that each target child made rapid increases in maintaining social interactions that generalized across people, settings, and materials and was maintained over a 2-month followup. There was less effect on social initiations.
Blew, Schwartz, and Luce (1985) used 1:1 peer tutoring involving two lower functioning elementary children and typical peers in community-based skills. A multistep adaptive skill sequence was task analyzed for each of four community settings (e.g., checking out a library book, making a purchase in a store). The typical peer was first taught discrete trial teaching techniques, and then taught the steps of each adaptive behavior sequence. After baseline (an ABAB design was used), the peer first modeled each step. Then he/she modeled and directly taught each step through instruction, prompting, and cuing, as needed. Both target students mastered at least two of the three sequences taught by a peer, but this required full peer teaching and reinforcements for both children in order for acquisition to occur. However, improved attention and responding of the child with autism to the peer was accomplished by the procedure.
Krantz and McClannahan (1993) examined the responses of four verbal children with autism to a scriptfading procedure described elsewhere (Krantz & McClannahan, 1998, p. 5). Printed cues in work schedules were used to stimulate social initiations to other peers with autism. A multiple baseline approach was used and was staggered across the four subjects. Each child increased the number of peer initiations dramatically, maintaining them when the cues were faded. Three of the four generalized across people, and all used novel language to initiate. The intervention also resulted in increased social interactions across all children including the untrained children.
- Rogers, SJ, Interventions that facilitate socialization in children with autism, Journal of Autism and Developmental Disorders, Oct. 2000 Vol. 30.
Reflection Exercise #12
The preceding section contained information
about interventions to facilitate socialization in school-age children with autism. Write
three case study examples regarding how you might use the content of this section
in your practice.
Online Continuing Education QUESTION 26
What were the results of daily exposure to typical peers who were motivated to interact with children with autism, but were untrained? Record the letter of the correct answer