Brief report: face configuration accuracy and processing speed among adults with high-functioning autism spectrum disorders.

Faja et al. (2009) asked 24 high-functioning adults with autism and 24 typical adults to spot small changes in face photos. The changes were either in the eyes, mouth, or spacing of features. Each person saw 96 pictures on a computer and pressed a key when they noticed a change. The team recorded accuracy and speed.

The task took 20 minutes. All volunteers had normal vision and an IQ above 80.

The autism group got 15 % fewer trials right. They only did as well as controls when the change was in the eyes. Reaction time was the same for both groups.

The authors say adults with ASD miss how faces fit together, not that they work slowly.

Hedley et al. (2015) ran a nearly identical lab set-up and also found worse face recognition in adults with ASD. They added upside-down faces and saw the same drop for both groups. Together the two papers show the deficit is real and not fixed by flipping the photo.

Becker et al. (2021) looks contradictory at first: they found adults with high autism traits judge neutral faces as more threatening, seeming faster, not slower. The difference is the task. Susan timed accurate feature detection; Casey timed gut threat ratings. Accuracy and bias are two separate channels.

Eussen et al. (2016) peered inside the brain. Adults with ASD showed slower amygdala habituation to fearful faces. The behavioral accuracy gap in Susan’s study lines up with the neural lag found here—both point to faces not being processed in a typical way.

When you teach social skills, do not assume clients caught the subtle eye narrowing or slight smile you showed. Use clear spoken cues or visual scripts instead of quick facial hints. Check understanding with a direct question, not a raised eyebrow. This small shift can cut confusion and boost compliance in high-functioning adults.