Both individuals with social anxiety disorder and individuals with autism spectrum disorder show alterations in the perception of others’ emotional facial expressions. Additionally, mimicry of an observed expression as well as the assessment of one’s own performance, i.e. metacognition, may be altered in these individuals. Using a non-clinical sample (N=57), we examined whether emotion recognition is linked to facial mimicry and confidence in one’s performance, as well as potential alterations in this link associated with social anxiety and autistic traits. While participants were presented with videos of spontaneous emotional facial expressions, we measured their facial muscle activity, and asked them to label the expressions and indicate their confidence levels in accurately labelling the expressions. Our results showed that confidence in emotion recognition was lower with higher social anxiety trait levels even though actual recognition was not related to social anxiety trait levels. Higher autistic trait levels, in contrast, were associated with worse recognition as well as a weakened link between performance and facial mimicry. Consequently, social anxiety might not affect emotion recognition itself but the top-down evaluation in an emotion recognition context, whereas individuals with autism may integrate information promoting emotion recognition, i.e. sensorimotor simulations, to a lesser degree.

Facial mimicry as well as the accurate assessment of one's performance when judging others’ emotional expressions have been suggested to inform successful emotion recognition. Differences in the integration of these two information sources might explain alterations in the perception of others’ emotions in individuals with Social Anxiety Disorder and individuals on the autism spectrum. Using a non-clinical sample (N = 57), we examined the role of social anxiety and autistic traits in the link between facial mimicry, or confidence in one’s performance, and emotion recognition. While participants were presented with videos of spontaneous emotional facial expressions, we measured their facial muscle activity, asked them to label the expressions and indicate their confidence in accurately labelling the expressions. Our results showed that confidence in emotion recognition was lower with higher social anxiety traits even though actual recognition was not related to social anxiety traits. Higher autistic traits, in contrast, were associated with worse recognition, and a weakened link between facial mimicry and performance. Consequently, high social anxiety traits might not affect emotion recognition itself, but the top-down evaluation of own abilities in emotion recognition contexts. High autistic traits, in contrast, may be related to lower integration of sensorimotor simulations, which promote emotion recognition.