Increasing social interactions of severely handicapped autistic children.
Teaching typical peers to start short, simple play invitations can pull even the most withdrawn autistic children into social contact.
01Research in Context
What this study did
The authors taught two typical classmates how to start play with two severely withdrawn autistic children. They used a multiple-baseline design across the autistic children. The peers learned simple openers like 'Come play' and 'Look at this.' Sessions happened during free-play time at school.
What they found
Both autistic children began talking and playing with the trained peers far more often. One child kept the new skills with new, untrained peers. The gains lasted after the coaches stopped checking.
How this fits with other research
This study is the middle link in a 50-year chain. Strain et al. (1977) first showed that teaching preschoolers to invite shy classmates works in regular classes. Castells et al. (1979) tried the same idea with autistic kids but saw no carry-over to new peers or new places. Barrett et al. (1987) trimmed the procedure and finally got one child to generalize. Laermans et al. (2025) later bundled peer training with teacher cues and doubled play time for four autistic preschoolers—an updated package that supersedes the 1987 lone-initiation tactic.
Why it matters
You do not need fancy toys or 1:1 adult shadowing. Pick one or two sociable classmates, give them a quick script, and set a play goal. Start with the target child you think will catch on fastest; the multiple-baseline look shows the effect spreads once peers stay consistent. If generalization stalls, add teacher prompts or move to the fuller Stay Play Talk model shown in Laermans et al. (2025).
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Choose one autistic client, pick two willing classmates, and spend 10 minutes teaching them to say 'Let's play' and hand over a toy—then watch free-play for five minutes and tally initiations.
02At a glance
03Original abstract
A peer-initiation training procedure was implemented across multiple peer trainers to investigate social interactions between severely withdrawn autistic children and their nonhandicapped peers. For one subject, substantial increases in spontaneous interactions with training and nontraining peers occurred after the peer-initiation procedure was applied across two training exemplars. Spontaneous social interactions continued even after the training procedure was removed. Although experimental control was established with the second subject during training, spontaneous interactions during nontraining periods were primarily with training peers. The results contribute to an emerging data base on the social interactions of autistic and severely withdrawn handicapped children and on peer-initiation training procedures.
Journal of autism and developmental disorders, 1987 · doi:10.1007/BF01487067