Loading the library…
autism interventions
The operative claimPeer-mediated interventions — training typically developing peers to model and prompt social interaction — reliably increase the social interaction of autistic students in inclusive settings and are among the better-supported social interventions, though effects are strongest for increasing interaction frequency and weaker for child-initiated social bids, with generalization and maintenance less consistently demonstrated.
Established that peer-mediated interventions increase autistic students' social interaction in inclusive settings; effects on spontaneous initiations and generalization are weaker.
Supported by systematic reviews and an overview of 10 reviews concluding PMIs are "strongly supported" and most robust in inclusive settings (Watkins et al., 2015; Reichow et al., 2024), a single-subject meta-analysis showing high effectiveness for social interaction (Zhang & Wheeler, 2011), and a What-Works-standards meta-analysis showing moderate-to-large effects on social communication (Watkins et al., 2019). A single-case review cautions effects are weaker for child initiations and that generalization evidence is limited (Martinez et al., 2019).
Watkins, L., O'Reilly, M., Kuhn, M., et al (2015). A Review of Peer-Mediated Social Interaction Interventions for Students with Autism in Inclusive Settings. Journal of Autism and Developmental Disorders.
Systematic review
Peer-mediated interventions are a promising, evidence-based approach for increasing social interaction in autistic students in inclusive settings, with positive generalization, maintenance, and social-validity outcomes.
Bearing on this claim: Systematic review: PMIs a promising evidence-based approach in inclusive settings.
doi.org/10.1007/s10803-014-2264-xZhang, J., & Wheeler, J. J (2011). A Meta-Analysis of Peer-Mediated Interventions for Young Children with Autism Spectrum Disorders. Education and Training in Autism and Developmental Disabilities.
Meta-analysis
Peer-mediated interventions were highly effective for promoting social interaction in young autistic children, with larger effects when peer modeling was used and maintenance/generalization were planned.
Bearing on this claim: Single-subject meta: highly effective for social interaction; modeling boosts effects.
doi.org/10.1177/215416471104600107Reichow, B., Lemons, C. J., et al (2024). Overview of Reviews of Peer-mediated Interventions for Young Children With Disabilities. Topics in Early Childhood Special Education.
Systematic review
Across all 10 reviews, PMIs improved social development of young children with disabilities (>80% of studies positive); the practice is "strongly supported" and should be considered an evidence-based practice, most robust in inclusive settings.
Bearing on this claim: Overview of 10 reviews: "strongly supported"; >80% positive; most robust in inclusion.
doi.org/10.1177/02711214241281383Watkins, L., Ledbetter-Cho, K., O'Reilly, M., et al (2019). Interventions for Students With Autism in Inclusive Settings: A Best-Evidence Synthesis and Meta-Analysis. Psychological Bulletin.
Meta-analysis
Interventions for autistic students in inclusive settings (including peer-mediated interventions) produced moderate-to-large effects, mostly on social communication; teacher-delivered interventions produced the largest effects.
Bearing on this claim: WWC-standards meta: moderate-large effects on social communication in inclusive settings.
doi.org/10.1037/bul0000190Counter-authority (1). Studies an adversary may cite — surfaced deliberately so the grade is honest about its limits.
Martinez, J. R., Werch, B. L., & Conroy, M. A (2019). Peer-Mediated Interventions to Address Social Competence Needs of Young Children With ASD: Systematic Review of Single-Case Research Design Studies. Topics in Early Childhood Special Education.
Systematic review
PMIs were effective for increasing interactions between autistic children and peers but may be less effective for increasing child-initiated social initiations or responses — and generalization/maintenance evidence is limited.
Bearing on this claim: Single-case review: weaker for child initiations; limited generalization/maintenance.
doi.org/10.1177/02711214198391364 supporting · 1 counter verified sources
Much evidence is single-case design rather than group RCT; effects are stronger for interaction frequency than spontaneous initiations; and generalization/maintenance are inconsistently demonstrated and depend on peer training and fidelity.
Last reviewed June 28, 2026