Impressions of Humanness for Android Robot may Represent an Endophenotype for Autism Spectrum Disorders.

Kumazaki et al. (2018) asked teens with autism and their parents to rate how human an android robot looked. A control group of typical teens and parents did the same task.

The team used a case-control design in a lab. They wanted to see if the way people judge humanness could act as a marker, or "endophenotype," for autism.

Both the teens with autism and their parents gave the android lower humanness scores than the control families did.

The pattern was strong enough that the authors suggest this rating gap could help subtype autism in future assessments.

Green et al. (2020) found that college students judge kids with visible stereotypy more harshly. Hirokazu’s work flips the lens: here, autistic observers are the judges, and they still differ from typical peers. Together, the studies show social perception is a two-way street in autism.

Griffith et al. (2012) review links repetitive behaviors to shared brain circuits. Hirokazu’s humanness-rating task may offer a simple lab proxy for those same circuits, giving clinicians a quicker signal than counting stereotypy.

Croteau et al. (2019) let autistic adults describe feeling labeled as "weird." The new data add a measurable twist: autistic people also read social cues (like robot humanness) differently, not just receive them differently.

You now have a five-minute, non-invasive probe—"How human does this robot look?"—that separates autism families from typical ones. No long checklist, no video coding. Try adding the android photo to your intake packet. If the parent and child both score it low, flag the chart for deeper social-cognition testing or use the score to track change during intervention.