Empathic Embarrassment Accuracy in Autism Spectrum Disorder.

Adler et al. (2015) asked adults with autism and matched controls to rate how embarrassed someone else felt in short video clips.

The team then compared each person’s ratings to the target person’s own embarrassment report.

This gave a clear accuracy score: how close the guess was to the real feeling.

Adults with autism were less accurate than controls at guessing the other person’s embarrassment.

They also gave themselves higher personal embarrassment scores, even when the situation did not call for it.

In short, they over-felt and under-guessed at the same time.

Hillier et al. (2002) saw a similar problem but found some spared skills, giving a “mixed” picture. The new study sharpens the lens by using exact accuracy scores, so the deficit now looks clearer.

Capio et al. (2013) showed that adults with autism feel normal vicarious embarrassment for accidents yet miss it when the act is intentional. Noga’s task removed intention clues, so the poor accuracy here lines up with M’s intentional-condition dip.

Trimmer et al. (2017) found that autistic bodies react to emotion films while the person reports feeling “less.” Noga flips the pattern: participants report “more” embarrassment while still mis-reading others. Together they show self-report can swing high or low, but other-awareness stays off target.

When clients say “I feel so embarrassed for them,” check if their rating matches what the other person actually feels. Train matching games: show a clip, pause, have the client predict the target’s 1–10 embarrassment, then reveal the answer and discuss the gap. Over time this can calibrate their social radar and cut down on awkward over-apologies or under-reactions.