Submitted by la334 on Thu, 01/10/2026 - 20:45
Cambridge researchers have brought together evidence from almost 2,000 participants to examine an influential theory of psychosis — finding a more complex picture than a general difference in how people use expectations.
Published in Psychological Medicine, the study was led by Chantal Miller-Silva and colleagues in the Department of Psychiatry, alongside international collaborators.
The team explored whether people with schizophrenia-spectrum psychosis rely differently on prior expectations when making sense of what they see and hear.
What we expect influences what we perceive
We make sense of the world using both the information coming through our senses and what we already know or expect. These expectations, known as “priors”, help us interpret information, particularly when it is unclear.
One influential theory suggests that an imbalance between expectations and sensory information could help explain experiences such as hallucinations and delusions.
But the evidence has been mixed. Some studies suggest that people with psychosis rely more heavily on expectations; others find that they rely on them less, or show no difference.
What did the researchers find?
The team combined 34 results from 27 studies, involving 904 adults with schizophrenia-spectrum psychosis and 1,039 healthy controls. Each study used tasks designed to test how expectations influence perception.
Taken together, the results showed no significant overall difference in reliance on expectations between the two groups. Separate analyses also found no significant overall association with hallucinations or delusions, although fewer studies were available for these comparisons.
The researchers also examined whether distinguishing between lower-level perceptual expectations and higher-level cognitive beliefs could explain the mixed findings. This distinction did not resolve the differences between studies.
A more nuanced picture
The findings do not rule out a role for predictive processing in psychosis. Some individual studies found differences in particular tasks, and results varied considerably across the research reviewed.
They do, however, suggest that findings from one task should be interpreted carefully before drawing wider conclusions about psychosis. The type of expectation being tested, the task itself, and how uncertain the sensory information is may all matter.
The authors recommend larger studies that use several different tasks with the same participants and, where possible, follow people over time and across episodes of illness.
This approach could help researchers build a clearer understanding of how expectations and perception interact in psychosis, and how that relationship connects to people’s symptoms.