My research focuses on learning and decision making. I approach these topics with a combination of behavioural experiments and computational modelling. I am also interested in optimal experimental design, in particular to distinguish between formal models of cognition.
In my research, I focus on how learning from others informs individual behaviour and decision making. That is, if we have evidence of other people’s behaviour without being able to query the reasoning behind the behaviour (e.g., observing recycling bins ready for collection outside someone’s house), how does that influence our own behaviour? - Do we recycle more, maybe because it appears to be a social norm? Recycle less because someone else is taking care of environmental problems, giving us the chance to free-ride? Infer the goal and apply it to a different area (reduce water consumption, for instance)? I am going to mainly use behavioural experiments and agent based modelling to answer these questions.
Ismail joined the lab when he did his MSc in Cognitive and Decision Sciences to work on research investigating learning transfer in the context of strategic interaction. After another MRes in Computer Science with a focus on Machine Learning, he has started his PhD in 2019, aiming to investigate the differential aspects of social learning in mental health disorders. He is interested in exploring whether behavioral markers developed through computational modelling of observed actions in online experiments can be used in diagnosing, monitoring and better understanding the determinants of social learning dysfunctions inherent to many disorders.
My MSc project investigates the impact of behavioural traces on green behaviour. More specifically, the project explores the following question: can people’s choice of elevator or stairs be influenced by showing them the choices made by others? This question will be examined by making otherwise invisible behavioural traces (i.e. choice of elevator or stairs) visible and testing their impact on behaviour.
My project investigates the role of transfer learning and generalization in human reinforcement learning. To what extent do humans employ knowledge of past, related environments to generalize and make inferences in novel environments? This is tested in a compositional multi-armed bandit task where the reward function is composed of previously learned functions. I also investigate how well kernel-based models describe human generalization.
Hannah joined the lab as a PhD student in 2013, and was jointly supervised by Chris Summerfield in Oxford. She worked on various projects in perceptual decision making, using behavioural studies, computational modelling, and EEG analysis. After completing her PhD, she took a job at Memrise and a part-time position as a post-doc in the Summerfield lab.
Eric first joined the lab as an MSc student in 2010, then as an MRes student in 2013, and finally as a PhD student (2014 - 2017). Eric worked on various projects, focusing in particular on functional generalization in reinforcement learning with Gaussian Processes. After leaving the lab, he became a Harvard Data Science fellow, working with Sam Gershman and Josh Tenenbaum. In 2020, he moved to Tuebingen to lead the Computational Principles of Intelligence Lab at the Max Planck Institute for Biological Cybernetics.
Yonatan joined the lab in 2019 for a research visit, sponsored by the Experimental Psychology Society. During his stay, he has been working on computational modeling of experience-based decision making.
Hrvoje first visited the lab in 2014 whilst a PhD student with Robin Hogarth at the Universitat Pompeu Fabra in Barcelona. He later returned for other visits in 2016 and 2017. We worked on various projects related to reinforcement learning, contextual bandits, and strategy selection. After receiving his PhD, he became a postdoc at UCL, working with Ray Dolan.
Adrian is a PhD student from the University of New South Wales. He visited the Speekenbrink lab in early 2018 to work on a project modelling reaction times in reinforcement learning tasks.
Charley is a PhD student at the Center for Adaptive Rationality at the Max Planck Institute for Human Development in Berlin. He visited the Speekenbrink lab in 2016 to develop the spatially correlated bandit paradigm.
Aline joined the lab from the MSc Behaviour Change programme to work on research investigating the effect of behavioural traces on donations to an environmental charity.
Alzbeta worked on an project investigating social learning in interpersonal games (the “beauty contest” game).
Felix joined the lab from the MSc Industrial/Organizational and Business Psychology programme to work on research investigating voice modulation and voice perception in a sales context.
Karolin joined the lab from the MSc Social Cognition to work on research investigating the effect of others’ judgments and their social group on updating beliefs by individuals.
Lucija joined the lab from the MSc Behaviour Change programme to work on research investigating the effect of behavioural traces on recycling behaviour.
Michael worked on a project assessing how an opponent’s strategy affects people’s own strategy in a simple two-player game.
Moritz worked on a project investigating planning in complex reinforcement learning tasks. After leaving the lab, he took on a PhD position at Warwick Business School, working with Nick Chater and Adam Sanborn.
Neo worked on a project investigating how Query Theory can explain the “pioneer effect”. After leaving the lab, he took on a PhD position at Warwick Business School.
Thomas worked on a project investigating the distinction between model-based and model-free reinforcement learning.
YS worked on a project investigating the distinction between state and reward functions in reinforcement learning.
Raphael worked on a project investigating reliance on model-free vs model-based reinforcement learning in a simplified two-player version of poker.
Anisha worked on a project investigating the relation between strategies and the evaluations of other players in a simplified two-player version of poker.
Jochen worked on a project investigating preference stabilization.
Thomas worked on a project related to the exploration/exploitation dilemma.
Anastasios worked on a project to investigate stabilization of preference with experience.
Audry worked on a project investigating model-based and model-free reinforcement learning.
Andrea worked on a project to investigate cue learning in a changing environment.
Akhat conducted his BSc Psychology research on how people learn about the volatility in the environment and how this affects their rate of learning.
Ryan worked on a project investigating framing effects in a multi-armed bandit task.
Anthony worked on a project investigating people’s ability to produce random sequences in games.
Benjamin worked on a project investigating preference stabilization.
Florence worked on a project investigaing the influence of volatility, stakes, and personality on exploration- exploitation behaviour.
James worked on a project investigating safe exploration in a reinforcemtn learning task.
Nadia worked on a project investigating the effect of information frequency on evidence weighting in perceptual decisions.
Nora worked on a project investigating the effects of foregone payoffs on exploration-exploitation behaviour.
Sofija worked on a project investigating how people learn to cooperate in multi-player games from their opponents’ strategies.
James worked on a project investigating the influence of information about foregone payoffs on exploration- exploitation behaviour.
Shouhao worked on a project investigating model-based and model-free learning.
Adam worked on a project investigating framing effects in decisions from experience.
Luke worked on a project related to model-based vs model-free reinforcement learning.
Zehra worked on a project to investigate whether there is a difference between learning when focussed on predicting vs. controlling outcomes.
Niroshan worked on a project investigating evidence integration in perceptual decisions.
Minho is an undergraduate student from South Korea, visiting the Speekenbrink lab at the start of 2018 as an intern to learn more about decision making and computational modeling.