Predictive processing theory suggests that perception occurs through the process of Bayesian inference. This involves the integration of sensory signals from the environment with pre-existing beliefs (‘priors’) about this input in order to determine the most likely cause of the sensory signal. In this view, perception does not simply result from passive bottom-up sensory processing but also involves the prediction of incoming stimuli based on our knowledge of the environment.
There is evidence that the use of priors in perception is altered in both patients with psychosis and non-clinical participants who have psychosis-like experiences (referred to as ‘schizotypy’). However, the literature has so far produced conflicting findings, and it is not known whether or not predictive processing is a unitary construct that underpins performance various cognitive domains. For example, does an individual’s tendency to rely on priors in one task predict their tendency to rely on priors in other cognitive domains, or is predictive processing domain specific? Our study aims to assess whether or not the tendency to rely on prior expectations is a construct that exists independent of particular cognitive domains, and whether these conflicting findings can be explained by the existence of different prior types at different levels of the cognitive hierarchy. We will then examine whether the use of these priors is altered in participants with high schizotypy scores indicating a proneness to delusions/hallucinations.

Student’s role

The student will assist with data collection in a large participant sample for this project, which involves administering a battery of tasks assessing prior expectations. The student will conduct data analysis/factor analysis and there is opportunity to gain experience in computational modelling techniques (computational psychiatry). The student will attend group meetings and seminars. The group is also active in studying clinical and basic cognitive neuroscience using fMRI. For further information/ notifications of interest please email cm954@cam.ac.uk (Chantal Miller, PhD student)