Observational learning by individuals with autism: a review of teaching strategies.

The authors read every paper they could find on teaching kids with autism to learn by watching others. They grouped the studies by the kind of modeling used: live models, video clips, or peer groups. They also noted when teachers added extra steps like asking the child to describe what they saw.

This is a narrative review, not a meta-analysis. They tell the story of the literature instead of crunching numbers.

Three tricks stood out. First, show a clear model on video or in person. Second, run the lesson with at least one other child so the learner can watch a peer. Third, teach the child to notice key parts of the scene with quick prompts or questions.

The review says these steps help kids with autism pick up play, social, and daily-living skills faster than trial-and-error teaching.

Hong et al. (2016) and Hong et al. (2015) later pooled single-case trials and confirmed video modeling is evidence-based for daily-living skills. Their meta-analysis gives the numbers the 2014 review lacked.

Diemer et al. (2023) looked deeper into the brain and found autistic people encode what they see just fine. The hitch comes later when they try to move their own muscles. This updates the review: modeling works because perception is intact, so focus follow-up training on motor practice, not clearer videos.

Older trials like Staats et al. (2000) and Lancioni et al. (2000) already showed video beats live modeling and that adding a verbal description step is key. The review stitches these threads into one simple package for clinicians.

You already use modeling, but are you pairing it with peer groups and a quick "What did you see?" check? The evidence chain says these small add-ons turn a good intervention into a great one. Start the week by running one social-skills group with a short video model and have each learner say the steps aloud before they practice. Measure initiation and see if you get the same jump the literature promises.