Gaze behavior: a new look at an old problem.

Wallander et al. (1983) looked at every eye-contact program for autistic kids. They asked: do the goals match how typical kids actually look at faces?

The team read all the papers they could find. They wrote a story-style review, not a lab test.

Most programs wanted kids to stare straight into eyes for many seconds. Typical kids rarely do this. They look at mouths, objects, then back to eyes.

The review warned: if we teach the wrong gaze pattern, we may set kids up to look odd, not natural.

McParland et al. (2021) tested a short classroom game that rewarded face-looking. Autistic and typical kids both doubled their looking time. This shows we can boost gaze without forcing long eye stares.

Wilson et al. (2019) tracked eye-move timing and found autistic kids disengage attention just as fast as peers. This supports the 1983 view: the problem is not a slow "unsticking" from eyes, but mismatched targets.

Rothwell et al. (2025) watched kids learn new words. Autistic children looked around more yet learned the words fine. Again, different gaze, same result—so rigid eye-contact goals may miss the point.

Before you write "maintain eye contact for 3 s" in a plan, check typical data. Break the goal into smaller parts: look at face, look away, look back. Reinforce flexible, natural shifts, not long stares. This keeps intervention friendly and avoids robot-like gaze.