Research News

New research combines computational modelling and neurophysiology to study decision-making

  • 08 December, 2020


A new research project will employ a powerful new approach to examine decision-making processes in the human brain.

The research addresses the questions: to what extent does most decision-making rely on the accumulation of ‘evidence’ over time; and what factors determine its extent? In resolving these questions, the project will generate new knowledge about the mechanisms of cognition in the brain, and new research tools that can be applied to study impaired information-processing in clinical conditions.

In the researcher’s new approach, neurophysiological signals reflecting decision-formation are traced at multiple levels of processing in the human brain and used to construct and constrain mathematical models of the decision process. This work will yield fundamental insights into core cognitive operations underlying flexible behaviour, revealing “how and why we adopt different decision algorithms in different situations.”

Wellcome Trust granted Associate Professor Simon Kelly from UCD School of Electrical and Electronic Engineering a prestigious Investigator Award in Science of over €800,000, the first award of its kind for University College Dublin.

He said: “We make countless simple perceptual decisions every day, any time we choose an action based on something we've seen or heard, for example. The algorithm your brain uses to make these decisions will be different depending on the demands of your environment – whether you're scanning the train platform to find your friend or choosing where to move among oncoming players in a rugby match.

“This investigator award is aimed at establishing how and why we adopt different decision algorithms in different situations. This problem has recently become tractable thanks to some methodological innovations that enable us to trace the evolution of a decision over time in electrical potentials measured non-invasively from the human scalp, and use them to construct mathematical models of behaviour.  

“This is important not just because it can shed light on mechanisms underlying a core cognitive function. There now exist many alternative mathematical decision models that are increasingly being used to understand various psychological phenomena, the functions of brain areas, and differences among individuals with and without mental illnesses. The trouble is, these alternative models can often be made to fit behavioural data equally well but lead to fundamentally different inferences.

“It is therefore critical to know whether, why and how these alternative mechanisms are invoked across different task scenarios, including ones representative of real life. By combining neurophysiology with computational modelling, this project stands to take major steps in this direction.”