Your eyes can deceive you: Predictive biases in perceptual representation of dynamic facial expressions.

It often feels like we can just look at someone's face and know how they feel. We sense when our coworker had a bad day or why our partner is smiling. Yet, although it may feel like we can simply read the expressions in others' faces, research suggests a more sophisticated process that uses what we know about the other person to "predict" how they should feel in this moment, and that what we seein their face reflects both: the expression they really show and the one we expect.
The proposed research tests this hypothesis, specifically whether prior emotion information about others - their body language, what they say about themselves - changes not only how we think they feel but also how we perceive their facial expressions. It will test, for example, whether a happy face will appear happier than it is with an excited body language, and less happy with a downcast body language. We will also investigate whether autistic individuals struggle in social interactions, paradoxically, because they see others' expressions more as they really are rather than how they should look like given the context.
This research will greatly enhance our understanding of how we read others' emotions and why we sometimes continue to misunderstand them, despite evidence to the contrary. The findings may have wide impacts, from emotion reading in law enforcement training to new interventions for autistic individuals, which currently only focus on a lexicon-like reading of emotions from facial
expressions.