Investigating reward-based motor performance in volatile environments using computational modelling and electroencephalography

Tecilla, Margherita. 2024. Investigating reward-based motor performance in volatile environments using computational modelling and electroencephalography. Doctoral thesis, Goldsmiths, University of London [Thesis]

Text (Investigating reward-based motor performance in volatile environments using computational modelling and electroencephalography)
PSY_thesis_TecillaM_2024.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (13MB) | Preview

Abstract or Description

Motor improvements have been linked to reward magnitude in deterministic contexts. Nevertheless, it remains unclear whether individual inferences about reward probability dynamically influence motor vigour. Moreover, how factors such as age, Parkinson’s disease or anxiety affect the modulation of motor vigour by predictions of reward probability remains unexplored.

This thesis, across four experiments, investigates how inferences about the volatile action-reward contingencies modulate motor performance on a trial-by-trial basis. We employed a reward-based motor decision-making task and modelled the behavioural data using the Hierarchical Gaussian Filter (HGF). In the final two studies, we also recorded the brain electrical activity through electroencephalography and used convolution models for oscillatory responses to delve into the neural underpinnings of motor decisions.

The results revealed that stronger predictions about action-reward probabilities led to faster performance tempo on a trial-by-trial basis in healthy participants. This effect was preserved in older adults and medicated Parkinson’s disease patients. Furthermore, the invigoration of motor responses extended to explicit beliefs (confidence) about reward tendencies.

Trait anxiety did not modulate the association between predictions and motor performance but affected practice effects over time. Analyses of the time-frequency representation of HGF computational quantities describing decision making unveiled increased alpha/beta correlates of different types of uncertainty among high trait anxiety individuals.

Finally, we found that state anxiety dampened the invigoration effect previously discussed. This manifested as longer reaction time for actions that were highly anticipated to yield rewards. Moreover, state anxiety led to reduced theta oscillatory responses during processing win/lose outcomes.

In conclusion, this thesis integrates computational modelling, Bayesian statistics, and electrophysiological approaches to explore motor decision-making behaviour under volatility. It provides novel evidence for an invigoration of motor performance by predictions about the action-reward contingency and sheds light on the modulation of this effect by age, Parkinson’s disease, trait and state anxiety.

Item Type:

Thesis (Doctoral)

Identification Number (DOI):


action-reward contingency; motor vigour; predictive coding; reward; uncertainty; EEG; Parkinson's disease; anxiety

Departments, Centres and Research Units:



31 March 2024

Item ID:


Date Deposited:

11 Apr 2024 11:30

Last Modified:

12 Apr 2024 08:32


View statistics for this item...

Edit Record Edit Record (login required)