Schizotypy, spanning nonclinical psychotic-like experiences to clinical psychosis, has been linked to disturbances of the self, particularly in the sense of agency (SoA). However, it remains unclear whether schizotypy reflects a bias toward self-attribution, reduced sensitivity to sensorimotor evidence, or both, and whether momentary SoA covaries with psychotic-like experiences.
We conducted a longitudinal Experience Sampling Method study in which 54 participants completed 60 online sessions. At each session, they reported their current psychological state, including SoA, and performed a validated agency attribution task involving parametrically distorted versions of drawings they had previously produced. We previously validated this task using behavioral modeling and EEG, showing that it captures retrospective SoA and its neural signatures. A hierarchical drift-diffusion model estimated self-attribution bias, sensitivity to sensory evidence, and sensitivity to performance feedback.
Higher schizotypy was associated with increased self-attribution bias and reduced sensitivity to sensory evidence. Transient reductions in SoA also coincided with decreased sensitivity to sensory evidence. Paranoia and positive schizotypy were additionally associated with altered self-serving bias and sensory evidence sensitivity. These findings reveal a dynamic relationship between schizotypy, momentary self-experience, and computational mechanisms supporting self-other distinction.
