Understanding the mechanisms behind deficits in imitation: do individuals with autism know 'what' to imitate and do they know 'how' to imitate?

Vanvuchelen et al. (2013) read every paper they could find on imitation in autism. They asked two simple questions. Do kids with autism know what to copy? Do they know how to copy it?

The team sorted the mess of results into two buckets. The first bucket is selection: picking the right action. The second is correspondence: moving your body to match it.

Both buckets leak. Some studies show kids pick the wrong action. Other studies show kids pick the right action but can’t do it. No one knows how the two problems mix.

The review says we need cleaner tests. We need to see if a child fails because she didn’t watch, or because she watched but her hands wouldn’t follow.

Whitehouse et al. (2014) give a sharp example. They split object movement from body movement. When the split is clear, kids with autism copy just as well as typical peers. This extends the review: the correspondence problem may vanish when the goal is obvious.

Gonsiorowski et al. (2016) look at toddlers who later get an ASD label. These toddlers look less at the demo. Less looking means less copying. This pins the selection problem on attention, not on motor skill.

Hermans et al. (2011) seem to disagree. They show kids can recognize a gesture yet still fail to copy it. This looks like a pure correspondence problem. But the tasks differ: Heidi used gestures, M et al. used objects. The clash fades once you see the method change.

When a child doesn’t imitate, first check what she watched. Put the toy under her eyes, not in your hands. If she still fails, break the action into tiny steps and give physical prompts. The review tells us to test both steps, not to jump to ‘poor motor planning’ or ‘lack of social motivation.’