Effects of peer-mediated social initiations and prompting/reinforcement procedures on the social behavior of autistic children.
Peer starts and adult prompts quickly boost social acts in autistic kids, but gains stay only while the program runs.
01Research in Context
What this study did
The team worked with autistic children in a classroom. They used two tactics at once.
Typical peers learned to walk over and start games. Adults also gave prompts and praise.
The design flipped back and forth four times to show true control.
What they found
Both tactics worked well during training. Positive social acts rose fast and stayed high.
When the program stopped, the gains vanished. Nothing carried over to free-play time.
How this fits with other research
Strain et al. (1977) tried peer starts with withdrawn preschoolers and saw the same quick lift.
Odom et al. (1986) added teacher prompts for autistic preschoolers and got longer play chains.
Syriopoulou-Delli et al. (2012) and Laermans et al. (2025) later cracked the generalization code. They taught peers in recess or Stay Play Talk and gains lasted after adults stepped out.
The 1979 paper is the base layer. Later studies kept the peer idea but tweaked delivery, setting, or added self-initiation drills to make play stick.
Why it matters
You now know peer starts give a fast bump, yet you must plan for transfer. Pair the peer package with recess practice, teacher priming, or self-monitoring so the child keeps playing when you leave.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Teach two typical classmates to invite the learner to play, then schedule recess follow-ups so the invites keep coming after you fade.
02At a glance
03Original abstract
Peer-mediated social initiations and prompting/reinforcement procedures were evaluated (in training and generalization sessions) as interventions for increasing the positive social behavior of four autistic children. During baseline, the peer trainer made few social initiations and did not prompt or socially reinforce subjects. For two subjects, baseline was followed by social initiation intervention, and for the other two, baseline was followed by prompting and social reinforcement. Both interventions produced dramatic and comparable increases in positive social behavior in training sessions. Post-treatment responding was not observed for either intervention. When interventions terminated in a second baseline period, the subjects' behavior returned to the level observed during the initial baseline. The subjects were then exposed to the intervention procedure they had not yet experienced. Again, there were positive and comparable behavior changes in the treatment setting, but no increase in positive social behavior was observed during generalization assessment.
Journal of autism and developmental disorders, 1979 · doi:10.1007/BF01531291