Social interaction and repetitive motor behaviors.

Matson et al. (2008) worked with three students with autism in a public school. Each student got the same three-part package: peers learned to invite them to play, the students learned to start conversations, and they tracked their own progress with a clicker.

The team used a multiple-baseline design across kids. They watched social initiations and hand-flapping, finger-waving, or body-rocking during recess and lunch.

All three students asked peers to play 4-6 times more often after training. Their repetitive motor behavior dropped to almost zero during the sessions.

One month later the gains were still there. Parents and teachers reported the same changes at home and in class.

Durbin et al. (2019) showed that simply letting autistic kids join regular music class boosts classmates' kindness. L et al. add a peer-training step, proving that teaching neurotypical kids how to invite builds even stronger social bridges.

Connell et al. (2004) used computer games to cut autistic children's echolalia. L et al. got the same stereotypy drop, but through real peer talk and self-monitoring, showing the goal can be reached with people instead of screens.

Groom-Sheddler et al. (2025) paired video modeling with self-management for poison safety. L et al. paired peer training with self-management for social gains. Both studies show self-monitoring is the shared engine that keeps skills alive after adult help fades.

You can copy this package tomorrow. Pick two friendly classmates, teach them to say "Come play tag," then teach the learner to ask first and click a counter each time. No extra staff, no tech budget. The study says one month of follow-up is enough to see if the routine sticks, so track it weekly and celebrate when the clicker stays quiet and the invites stay high.