Teaching Children with Autism in Small Groups with Students Who are At-Risk for Academic Problems: Effects on Academic and Social Behaviors.

Cox et al. (2015) mixed one child with autism and two at-risk readers into a single reading group. The teacher ran 15-minute lessons three times a week. Each lesson had two parts: direct instruction on letter sounds and planned turn-taking chats.

The team tracked the autistic child's correct academic responses and social bids during class. They also watched the same child at recess to see if the skills carried over.

Every student, including the child with autism, hit the academic goal—mastering four new letter-sound pairs. The autistic child also started raising his hand, sharing materials, and commenting far more often during lessons.

Recess data were hit-or-miss. Social gains showed up some days and vanished others. The authors call generalization 'inconsistent.'

Menezes et al. (2021) reviewed 18 similar studies. Their big picture: inclusive social-skills lessons work when peers take part. Cox et al. (2015) is one concrete example that their synthesis captures.

Dudley et al. (2019) looked at 27 school studies and found most social programs are run by researchers, not teachers. Cox et al. (2015) pushes back by showing a classroom teacher can run the group alone and still win.

Alwahbi et al. (2021) tried peer training first; it did nothing until they added a contingency contract. Cox et al. (2015) skipped contracts and still got social gains, suggesting that built-in academic turn-taking can be enough if the activity is structured right.

You can fold a child with autism into an existing academic group without pulling anyone out. Use short, scripted lessons that build in social turns—like passing cards or choral responses. Track both reading sounds and social bids; both can climb together. If recess carry-over is weak, add a short peer reminder just before lunch.