Signal detection and false perceptions: Investigating the integration of bottom-up and top-down processes in auditory hallucinations

Abstract

Auditory verbal hallucinations are the sensory experience of hearing a spoken voice in the absence of an external speaker. Currently, there is no satisfactory explanation as to the cognitive mechanisms underlying these phenomena. To better understand why they occur, researchers often use experimental paradigms to assess perceptual errors, including the Auditory Signal Detection Paradigm. However, inconsistencies in the reported strengths of task effects, and differences in the design of Auditory Verbal Signal Detection Tasks (AV-SDTs), have made generalisation of results challenging. This thesis examines the historical use of AV-SDTs in hallucinations research and explores how this paradigm can be used to investigate the construction of perception and the occurrence of misperceptions. A meta-analytic review of AV-SDTs confirmed that liberal response biases are associated with hallucinations. Additionally, the review revealed perceptual sensitivity deficits associated with the presence of hallucinations in non-clinical cohorts. Following this, an experimental study validated a standardised AV-SDT for online use. Using this design as a baseline, a series of experimental studies manipulated the task design to assess how strengthening different types of perceptual expectations influenced perceptual judgements and beliefs in non-clinical samples. Two studies showed that strengthening cognitive expectations led to changes in response behaviour and beliefs but no changes in perceptual ability. Finally, two studies then showed that the way perception is shaped by sensory expectations is dependent upon the context in which the subject hears auditory signals. Collectively these findings highlight that the extent to which the brain relies on top-down versus bottom-up processes differs depending upon the context and the type of information it receives.

Date of Award	30 Apr 2025
Original language	English
Awarding Institution	Northumbria University
Supervisor	Emma Barkus (Supervisor) & Peter Moseley (Supervisor)