Agent-based Modelling for the Social Scientist
4th - 6th February 2019, University of Surrey, Guildford, UK
Simulating social interactions in virtual research labs using agent-based modelling is increasingly allowing researchers to gain new insights into the complex ways that individuals and societies function.
This 3-Day course, provided by the Department of Sociology at the University of Surrey, will cover the process of agent-based modelling, from conceptualising a research question, where to obtain data, operationalisation and formalisation of data, model implementation, and model analysis and interpretation. In addition to the theoretical content, you will learn NetLogo as a programming language for agent-based models. On the basis of a detailed model of a social phenomenon (protective behaviour when faced with an epidemic) that is developed step-by-step in lab sessions, the major features of programming in NetLogo are learned through practical application. Through this guided implementation you will acquire basic to intermediate programming skills in NetLogo as well as engaging with the step-by-step development of a model.
Indicative content includes:
- What is agent-based modelling
- Basics of agent-based model implementations
- Approaches to behaviour rules (eg. game theory, BDI, social psychology)
- Running and analysing experiments
- Sensitivity analysis and robustness tests
- Verification and validation
On successful completion of this course, particpants will be able to:
- Understand the foundations of agent-based modelling (K)
- Understand application areas of agent-based modelling (C,K)
- Understand different implementations of social phenomena (C,K)
- Be able to program in NetLogo (K,T,P)
- Be able to provide a basic model specification and a basic implementation (P)
Key: C-Cognitive/Analytical; K-Subject Knowledge; T-Transferable Skills; P- Professional/ Practical skills
Dr Corinna Elsenbroich is computational social scientist. Her main research interests are in methods development, in particular methods for complexity social science and methodological and epistemological aspects of agent-based modelling and social simulation. She is interested in understanding decision mechanisms, in particular collective decision-making and context dependency of decisions.
Dr Jennifer Badham originally trained as a mathematician and developed an interest in applying mathematical modelling methods to social policy while working for government and nongovernment health organisations in Australia. Her main research interest concerns the way that social structures affect transmission – of disease, information, beliefs and behaviour. This brings together aspects of social simulation, social network analysis and social psychology. She is currently working at Queen's University Belfast modelling health behaviour interventions over social networks.
Level of Study:
Entry (no or almost no prior knowledge)
Varies according to status:
- £595 - Government/commercial sector
- £495 - Educational/charitable sector
- £395 - Students.
Gilbert, N. (2008). Agent-Based Models. Number 153 in Quantitative Applications in the Social Sciences. Sage Publications, 2008.
Gilbert, N. and Troitzsch, K. (2005) Simulation for the Social Scientist, OUP.
Squazzoni, F., Jager W. and Edmonds B. (2014) Social Simulation: A Brief Overview. Social Science Computer Review, 32 (3).
We suggest you bring your laptop. If you would like to use the university computers a visitors login will be provided and the software will be available. In that case, please bring a USB stick to enable you to take the model developed during the course home with you.
Participants on the course will include some students completing the MSc in Social Research Methods
How to book - Reserve your place by registering and paying via the University of Surrey Online Shop.