Dr Jens Madsen

This course is available on the MSc in Behavioural Science, MSc in Organisational and Social Psychology, MSc in Social and Cultural Psychology, MSc in Social and Public Communication and MSc in Societal and Environmental Psychology. This course is freely available as an outside option to students on other programmes where regulations permit. It does not require permission. This course uses controlled access as part of the course selection process. For information on controlled access courses, including eligibility, application processes, deadlines, and departmental contact details, please refer to the Controlled Access Courses webpage.

How to apply: All PBS 0.5-unit courses in Winter Term are controlled access and capped. Students enrolled on PBS programmes will be given priority.

Each course is available with permission as an outside option to students outside of PBS where regulations permit, providing there is space. All students must submit a short statement (around 100 words) outlining their motivation for enrolling on the course, which will be considered by the course convenor.

Deadline for application: Please apply as soon as possible after the opening of course selection for all courses.

For queries contact: Pbs.msc@lse.ac.uk

Recommended pre-requisites:

The Module is computational in nature and will use formal coding (R, Python, etc.). We recommend that students have some experience with R (or a similar language)

Additional requisites:

There is no pre-requisite requirements for taking the course, as the Harvard-Style seminars will give students the needed skills to complete the assignments and reach the Learning Outcomes. Of course, given the quantitative nature of the course, it is helpful is students have taken introductory quantitative classes such as MY451 or similar courses. In addition, while it is not a necessary requirement, it would be useful if students have used coding programs like R to conduct statistical analyses. 

The first half of the course focuses on individual-level computational approaches. We discuss the philosophical and scientific methodological assumptions that underpin the approaches to individual behaviour and exemplify this via Bayesian modelling. At the end of the first half, students should be able to generate their own Bayesian inference models of psychological and behavioural assumptions.

The second half focuses on meso-level computational approaches. That is, behaviours that unfold in dynamic contexts where people decide what to do within socio-cultural, economic, and organisational constraints and where the actions of others can influence their behaviour. These interactions can cause feedback loops in the system, which makes system complex and malleable. We exemplify this via agent-based modelling. By the end of the second half, students should be able to generate the structure of a simple agent-based model.

In summary, the course is relevant to psychologist or behavioural students – as well as social science students more generally – who want to enhance their analytical and technical skills and to foster a deeper understanding of human behaviour through the lens of computational science. The following is a proposed overview of the course structure across a 10-week period. While the specific course content will be finalised later, the following provides an example of the lectures and classes that would take place during the module. 

Week

Theme

Lecture

Class

1

Introducing computational modelling

Discuss the history and philosophy underpinning computational modelling

Foundational principles

2

An overview of approaches to minds and societies

Different computational approaches to formal representations

Discussion of different modelling approaches

3

Bayesian modelling 1

Bayesian theorem and inferences

Bayesian reasoning as a case study; forward and backward updating

4

Bayesian modelling 2

Operationalising Bayesian Networks and making predictions

Introducing GeNie and tools for Bayesian modelling

5

Testing individual-oriented computational models

Comparing Bayesian networks with empirical findings and model comparisons

Link between data sciences and individual-oriented models

6

READING WEEK

READING WEEK

READING WEEK

7

Complex societal problems

Numerical tools to capturing complex systems

Introducing Netlogo and tools for agent-based modelling

8

Computing psychological and behavioural assumptions

Inputting social and cognitive psychological and behavioural assumptions into ABMs

Two example models – one information theoretic; one on COVID transmission

9

Calibrating and validating agent-based models and stress-testing systems

Comparing ABMs with empirical findings and discussing fragility of complex systems

The relationship between data science and computational modelling

10

An agent-based model from concept to application

Showcasing cyclical nature of developing and testing an ABM as well as behavioural model comparisons

Sustainability models and calibration

11

Challenging models of minds in societies

Limitations to concepts, testing, and assumptions.

Critical discussions of assumptions

The course will give students an introduction to the philosophical assumptions that underpin computational models in psychological and behavioural science as well as allow students to do a deeper dive into two specific methods (Bayesian and agent-based models) and how these relate to psychological and behavioural science. We will look at misinformation and environmental sustainability as examples of the models, but the course will discuss general principles of how to translate psychological and behavioural theories into computational models.

15 hours of lectures and 20 hours of seminars in the Winter Term.

This course has a reading week in Week 6 of Winter Term.

An outline of a novel computational model to describe a psychological or behavioural scientific theory or idea. This will be developed and refined in seminars with feedback from the course convenor and GTA.

Indicative Reading:

Pearl, J. (2009). Causality. Cambridge university press.

Epstein, J. & Axtell, R. (1997) Growing Artificial Societies: Social Science from the Bottom Up, MIT Press

Griffiths, T. L., Chater, N., & Tenenbaum, J. B. (Eds.). (2024). Bayesian models of cognition: reverse engineering the mind. MIT Press.

Johnson, N. (2007) Simply Complexity: A clear guide to complexity theory, One World

Lee, M. D., & Wagenmakers, E. J. (2014). Bayesian cognitive modeling: A practical course. Cambridge university press.

Miller, J. H. & Page, S. E. (2007) Complex Adaptive Systems: An Introduction to Computational Models of Social Life, Princeton University Press

Oaksford, M., & Chater, N. (2007). Bayesian rationality: The probabilistic approach to human reasoning. Oxford University Press, USA.

Schelling, T. C. (2006) Micromotives and Macrobehavior, W. W. Norton & Company

Sun, R. (2006) Cognition and multi-agent interaction: From cognitive modelling to social simulation, Cambridge University Press

Project (100%).

A novel computational model that translates a psychological or behavioural scientific theory or idea into a coded model (R, Python, NetLogo, or similar), accompanied by a 750-word explanatory document that describes the motivation, functions, assumptions, and limitations of the model.

The assessment emphasises creativity in thinking - the ability to identify an interesting idea, translate it into a structured model, and justify the choices made. The model does not need to be calibrated against real data, but it should demonstrate theoretical grounding, sensible behaviour, and awareness of the modelling choices involved.