Evaluation of the high inflexible precision of prediction errors in autism theory using simple and biological motion paradigms

Todorova, Greta Krasimirova (2021) Evaluation of the high inflexible precision of prediction errors in autism theory using simple and biological motion paradigms. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2021Todorovaphd.pdf] PDF
Download (19MB)


Autism Spectrum Disorder is a lifelong neurodevelopmental condition associated with large lifetime costs to services, the individuals and their families. To develop appropriate support, a unified account of the condition needs to be found to provide a framework for research to build on. This thesis focuses on testing one of the recently proposed theories – the theory of High Inflexible Precision of Prediction Errors in Autism (HIPPEA), which interprets autism through a predictive coding perspective. The predictive coding framework argues that through experience, the brain forms predictions about the incoming sensory information, which it then compares with the actual input. Mismatches produce prediction errors which are weighed in comparison to the prediction, and if enough weight is assigned to the precision of the prediction error, a change in the prediction or the action is enacted. HIPPEA poses that autism arises from a difference in the tuning of this general neurocognitive mechanism whereby attention leads to the invariably high precision setting of prediction errors. This, in turn, leads to the creation of narrow prediction models that are based on infrequent contingencies and noise. This PhD aims to contribute research results and paradigm designs that investigate precision weight setting of prediction errors in autism.

This thesis presents three behavioural and one neuroimaging experiments, and one meta-analysis. Each experiment modulates attention and expectation under different experimental paradigms allowing for the investigation of these two factors in multiple contexts. Chapter 2 makes use of an established apparent motion paradigm. In this chapter, endogenous attention is controlled allowing the investigation of the differences in prediction establishment and prediction error processing in neurotypical and autistic individuals. Moving forward, to establish the viability of using biological motion stimuli as an effective way to measure differences between autistic and non-autistic individuals, Chapter 3 presents a large-scale meta-analysis of behavioural, eye-tracking, EEG and fMRI studies investigating biological motion perception and interpretation in autism. Chapter 4 presents two studies that look at the effects of autistic traits in a task that orthogonally modulates attention and expectation by explicitly instructing participants about the statistical regularity of events and by providing implicit cuing using a human point-light kicker or a coherent dot-motion display. Finally, Chapter 5 presents a proof-of-concept study, which examines the feasibility of a modification in a recently developed EEG paradigm of hierarchical frequency tagging of bottom-up and top-down signals using dynamic human biological motion. This paradigm allows the investigation of the representation of low- and high-level components of the human point-light display, along with their integration in the brain while modulating attention and expectation through task instruction.

The results from this thesis indicate that like neurotypical participants, autistic individuals can create and benefit from the development of predictions either through illusory motion or through the explicit establishment of expectations. In line with HIPPEA, this indicates that it is not the establishment of predictions that is the cause of the traits observed in autism. Moreover, what we see is that unpredictable events are treated differently, suggesting disproportionate amplification of unpredictable events, as suggested by HIPPEA. However, we do not see support for the ‘inflexible’ part of the HIPPEA theory. Instead, this thesis concludes that prediction errors show some special treatment in autism, but that is context-dependent. For research to move forward, it is paramount that attention is a controlled factor, and that context-dependent precision weight setting of prediction errors is incorporated in a reviewed version of the HIPPEA theory.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Keywords: autism spectrum disorders, biological motion, meta-analysis, age, emotion recognition.
Subjects: B Philosophy. Psychology. Religion > BF Psychology
Colleges/Schools: College of Science and Engineering > School of Psychology
Supervisor's Name: Pollick, Professor Frank and Muckli, Professor Lars
Date of Award: 2021
Depositing User: Theses Team
Unique ID: glathesis:2021-82527
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 22 Oct 2021 10:21
Last Modified: 27 Mar 2023 10:47
Thesis DOI: 10.5525/gla.thesis.82527
URI: https://theses.gla.ac.uk/id/eprint/82527
Related URLs:

Actions (login required)

View Item View Item


Downloads per month over past year