A comparison of peer-initiation and teacher-antecedent interventions for promoting reciprocal social interaction of autistic preschoolers.

The team worked with autistic preschoolers during free-play time.

They compared two setups head-to-head on alternating days.

One setup taught typical peers to walk up and start play.

The other setup had teachers give quick prompts before play began.

Each child experienced both setups many times so results could be fairly compared.

When peers started the play, autistic children answered back more often.

When teachers gave quick prompts first, the same children both started play and answered more.

Teacher-first setup also created longer back-and-forth play chains.

Both ways helped, but teacher-antecedent gave fuller social pictures.

Barrett et al. (1987) ran a near-copy of the peer-initiation arm and saw the same jump in responses, showing the effect holds across kids and classrooms.

Fingeret et al. (1985) mixed peer starts with teacher prompts and tokens; they warned that interaction dropped once the teacher stopped cueing, hinting that pure peer work may need booster help.

Koegel et al. (1992) later showed you can fade those teacher prompts and still keep the gains, solving the maintenance worry raised by the 1985 paper.

Laermans et al. (2025) bundled teacher facilitation into a tidy Stay-Play-Talk package, proving the teacher-antecedent idea is still alive and scalable today.

You now have a quick decision tree for social-skills blocks.

Need more responses right now? Train three typical peers to approach and invite play.

Need balanced initiations and longer chats? Add brief teacher prompts before free play and coach peers to keep the ball rolling.

Run each setup for a week, graph turns and chains, and keep the one that gives you the fuller dance floor.