Comparing the teaching interaction procedure to social stories: a replication study.

Three autistic kids tried two ways to learn social skills. One way was the teaching interaction procedure: the adult asked a quick question, showed the skill, then the child practiced with feedback. The other way was reading a short social story.

Each child got both treatments in random order. The team counted how many social steps the child did correctly during play with a peer.

Teaching interaction won every time. All three kids reached a large share correct steps after only a few sessions. Social stories helped a little, but scores stayed lower and slower. No teaching at all produced almost zero progress.

Cramm et al. (2009) had earlier shown that social stories can work. The difference: their kids already knew the game, so the story only cued what to do. Alyne’s kids were learning brand-new skills, so they needed the full demo and practice loop.

Camilleri et al. (2024) and Lde Leeuw et al. (2024) now show social stories still have life when parents or apps deliver them digitally to younger verbal kids. Alyne’s lab result does not cancel those gains; it just sets the bar higher when you want fast, full learning.

Morgan et al. (2014) ran a twin study the same year: adults with autism got the same behavioral-skills-training package for job interviews and also beat a wait-list. Same method, new age—showing the teaching interaction recipe travels well.

If you need rapid skill gain in school or clinic, choose teaching interaction. Ask, show, and let the learner rehearse with quick praise and corrections. Save social stories for reminders after the skill is mostly there, or for kids who just need a cue. You can still hand the story job to parents or an app later—just don’t expect it to teach the whole skill from zero.