Bayesian analyses of the effect of metacognitive training on social cognition deficits and overconfidence in errors

Social cognition (SC) impairments in schizophrenia predict interpersonal problems and low functional outcome, which might be aggravated by low (meta-) cognitive awareness of individual symptoms and cognitive biases. Metacognitive Training (MCT) aims to raise patients' awareness of cognitive biases, for example overconfidence in errors. We examined whether MCT reduces high-confident false mental state perceptions and tried to identify possible underlying mechanisms of SC impairments. A total of 150 patients were enrolled in a randomized clinical trial comparing the MCT with cognitive remediation (CogPack®) as the active control. Participants were assessed at baseline and at four weeks (post) and further six months (follow-up) later with the Reading the Eyes in the Mind-test also measuring the patients' response confidence. We found that compared to CogPack® MCT reliably reduced the amount of overconfident SC errors by approximately 40% at follow-up. Additionally, we were able to link several symptomatic features and neuropsychological parameters to SC impairments and overconfidence herein.