Can we predict general social functioning from patterns of conversational interaction in clinical populations? How do interactional patterns relate to more traditional measures of social cognition and neurocognitive skills? This project collects diagnostic, neuro-cognitive, voice and movement data in 40 patients with schizophrenia or schizo-affective disorder and 40 matched controls. The dataset analyzed through cutting edge machine learning methods will be used to understand the potential of interactional data to assess symptoms and functional outcome. Not least we will investigate how social impairment is grounded in distinctive behaviors and reaction patterns during social interactions.