Understanding of others' intentions in children with autism.

The team watched preschoolers with autism and peers with developmental delay. Each child saw an adult try but fail to complete an action, like pulling a toy apart. Kids then got the toy. Researchers asked: Do they copy the goal the adult never finished?

No drills, no rewards—just see and do. The set-up tested if autistic kids read unspoken intentions.

Both groups copied the missed goal the same amount. Autistic children understood the adult wanted to finish even though it never happened.

In short, intention-reading was not weaker in autism once language level was matched.

Somogyi et al. (2013) looked at low-functioning, non-verbal children with autism. Those kids copied odd hand motions exactly, showing they saw the goal but missed the why. Richman et al. (2001) found no such gap in preschoolers who could talk, so the two studies together draw a line: language level changes what looks like an intention problem.

Du et al. (2024) used fancy headbands to watch brains while kids imitated. They still saw looser, slower copying in autism. Their tech confirms the behavior Richman et al. (2001) captured, but shows the gap widens when movements get more complex.

Demily et al. (2018) tested bright young adults with autism. The adults imitated perfectly when told to, yet failed to plan comfy, efficient moves on their own. Again, imitation can look fine in structured tasks—just like Richman et al. (2001) showed—while spontaneous use stays shaky.

Do not assume a child with autism cannot read your intentions. If their language matches their age, they likely know what you meant to do. Use that strength: model goals clearly, then let them finish the action. When language is low, shift to simpler goal cues and repeat; intention hints may go unseen. Keep tasks short and reinforce each success—complex or free-form settings will still trip them up.