I am currently a Postdoctoral Research Fellow at the Max Planck UCL Centre for Computational Psychiatry and Ageing Research at University College London, where I work with Professor Quentin Huys. My research interests revolve broadly around the interplay between emotions and decisions, and what this means for people with depression.
You can read more or contact me on my personal website.
I did my PhD in the MBB team between 2018 and 2022 with Mathias Pessiglione. I was affiliated with the Ecole Doctorale 3C and received funding for the 4th year of PhD from the Fondation pour la Recherche Médicale.
During my PhD, I researched how incidental affective states influence economic decision-making. The studies I conducted involved emotions and mood fluctuations that are incidental to the decision at hand - so that any spillover that might occur is 'irrational', according to the maxims of behavioural economics. We asked whether we can observe and measure such a spillover, and if so, whether it is truly irrational. Below is a breakdown of the projects I was involved in - but you can also hear me present them in under 45 minutes during my PhD defense.
We are all familiar with the phenomenon of varying moods as the seasons change. This was no different for our evolutionary ancestors, because of seasonal variations in the availability of different sources of food. Typically, the abundance of a food sources are (auto)correlated. This property of the environment implies that the decision to forage and informs an impression of the momentary richness of the environment - we call this 'mood'. A positive mood then stimulates the agent to forage more, whereas a negative mood biases the foraging decision towards staying put. A small mood bias therefore helps the agent in making an informed decision to explore for food. However, our simulations showed that a bias that is too large can exaggerate this propensity to explore when the environment is rich, resembling manic behaviour. And, by consequence, once the environment gets poorer, the agent's willingness to forage plummets into a prolonged depression. With this theoretical project, we have argued how a mood bias can be adaptive, thereby giving an explanation for why mood fluctuations have been preserved throughout evolution. We also showed that an extreme mood bias can explain behaviour that resembles what we observe in mood disorders.
The propensity to seek out costly but rewarding options when we are in a happy mood (versus preferring less rewarding but uncostly alternatives when we are in a sad mood) can be captured as a decision bias parameter when we model decision-making during economic cost-benefit trade-offs. In a battery of choices, participants in the lab get to choose between accepting a risk, delay, or physical/mental effort in exchange for a large sum of money, or alternatively, to get a lower sum of money for free. I developed an algorithm that is designed to optimise such an economic decision-making experiment in realtime, so that no time-consuming calibration is necessary because the task adapts to the player. Participants performed these economic choice trials interleaved with Fabien Vinckier's quiz task, where phases of the game featuring lots of positive feedback were alternated with phases where subjects got many questions wrong. The ensuing mood fluctuations are incidental to the economic decisions, but they indeed spilled over.
When we induced short-lived emotions (as opposed to slow mood fluctuations in the above project) with short text scenarios paired with congruent music, participants not only rated that they felt happy or sad, but also displayed emotion responses in their physiological reactions. Valence signals (smiling and frowning muscles of the face) and arousal signals (pupil dilation, skin conductance) could be combined to a physiological mood proxy that significantly correlated with preference fluctuations towards rewarding but costly options when participants were happy, versus uncostly and less rewarding options when they were sad. Moreover, we could observe that this bias emerges early on in the decision process by using eyetracking. We therefore concluded that the mood bias can indeed be interpreted as a predisposition.
This paper is currently under review. Stay tuned!
It has long been known that people gamble more when the weather is nice or the local sports team just celebrated a victory. We wondered whether we can again observe a propensity to seek rewarding but costly decisions when people's mood is positive (and vice versa when negative) even outside the lab, in the absence of any induction whatsoever. We thus conducted an online experiment where a cohort of subjects logged in weekly to rate their mood, and performed a short battery of economic cost-benefit trade-offs. We found that indeed, the decision bias caused by real-word mood fluctuations was significant. Moreover, this bias was correlated with the bias that we observed in the same people at the very beginning of the study, when they did an intake experiment that featured the same emotion-induction paradigm as in project [3] above. Thus, people's real-world mood biases can be predicted from a short experiment!
This paper is currently forthcoming. Stay tuned!
The method (described in project [2]) to optimize the experimental design in real-time for economic cost-benefit trade-off tasks can be universally applied - it is not specific to mood experiments! In ongoing collaborations, others have used the algorithm in studies that involve mental fatigue, physical fatigue, or antidepressant drug treatment. If you are interested in this algorithm, have a look at this wiki on the MBB Github.
My emotion induction stimuli (text vignettes and music fragments) were successful at inducing happiness, sadness, anger, fear, or a neutral baseline state. If you are interested in using my stimuli, drop me a line.