Use this URL to cite or link to this record in EThOS:
Title: Bayesian hierarchical predictive coding of human social behaviour
Author: Hillebrandt, H. F.
ISNI:       0000 0004 5358 2876
Awarding Body: University College London (University of London)
Current Institution: University College London (University of London)
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
‘Bayesian hierarchical predictive coding of human social behaviour.’ Biological agents are the most complex systems humans encounter in their natural environment and it is critical to model other’s mental states correctly to predict their behaviour. To do this one has to generate a mental representation based on an internal neural model of the other agent (Chapter 1). Here we show, in a series experiments, that people use and update their Bayesian priors in social situations and explain how they create mental representations of others to guide action selection. We investigate the neural mechanisms and the brain connectivity that underlie these social processes and how they develop with age. In chapter 2, we show how experimentally induced prior experience with other people (here social inclusion or exclusion) influences the level of trust towards those people. In chapter 3, we describe an fMRI study using a social perspective-taking task that examines the developmental differences between adolescents and adults in the control of action selection by social information. Using the same task, in chapter 4, we investigate the effective connectivity between the activated regions with Dynamic causal modelling. In Chapter 5, we explore effective connectivity of fMRI data from the Human connectome project (Van Essen et al., 2012). During the task participants viewed animations of triangles moving either randomly or so that they evoke mental state attribution (Castelli et al., 2000). Chapter 6 concludes with a summary of the experiments and integrate them into existing research, as well as provide a critical synthesis of the findings in order to suggest future research directions. We interpret our findings in a hierarchical predictive coding framework, where agents try to create a neural model of the external world to minimize prediction errors, Bayesian surprise and free energy.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available