Social cognition (SC) is an impaired domain in schizophrenia. However, little is known on the Signal Detection properties of SC deficits.

We analyzed the relationship between emotion perception and psychotic symptoms in a sample of schizophrenic patients. For this scope, we extended the scoring system of the awareness of social inference task-emotion recognition (TASIT-ER) according to signal detection theory (SDT).

Sample:

– one hundred and nineteen inpatients from L’Aquila Inpatient unit diagnosed with schizophrenia.

Dependent variable:

– Positive and Negative Syndrome Scale (PANSS)’s Positive, Negative, Disorganized, Excited and Depressed dimensions, and total score.

Independent variable:

– a modified version of TASIT-ER. The original scoring system, including only “HITs”, was extended with “False Alarm” (FA), defined as a detection of an emotion when not present.

Statistical analysis:

– multivariable linear regression models for each sub-group of emotions to assess the effect of FAs on psychotic symptoms compared to HITs.

FAs on positive emotions were associated with disorganized (b = 31.95), excited dimensions (b = 41.84) and PANSS Total (b = 152.46); FAs on negative emotions were associated with Excited dimension (b = −57.97) and PANSS Total (b = −243.70). HITs on Negative emotions were associated with Negative (b = −13.37), Disorganized (b = −8.64) Excited (b = −8.74) dimensions and with PANSS total (b = −45.30).

FA rates were more strongly associated with total PANSS score than HIT rate, suggesting a prominent role of false recognition in defining psychotic symptoms, especially disorganized and excited ones, consistently with computational models of psychosis that rationalize false recognition as failures of active inference systems in updating their predictive model of sensory information.

The authors have not supplied their declaration of competing interest.