Agent-based models in social science

ESRC Methods North West training and networking workshop

Social science research that is interested in the spread of communication or behaviour is often restricted by how much we can observe, whether that’s via behavioural experiments, social media data or ethnographic approaches. Agent based models (ABMs) can help us investigate our observations and hypotheses by combining theory with observations to carry out virtual experiments, which can be scaled up and used to explore different scenarios.

ABMs are relevant to many areas of research that involve things interacting. We will introduce concepts important to ABMs in social science, including:

• Social behaviours, networks and interactions
• Example applications and the role of ABMs
• Building ABMs with user-friendly software
• Using experiments to support and validate models

There will be talks from:

• Dr Corinna Elsenbroich. Reader of Computational Modelling in Social and Public Health Science, University of Glasgow.
• Dr Christine Cuskley. Research Scientist and Consultant at CogKnit; formerly Reader in Language and Cognition, University of Newcastle.
• Dr Hannah Little. Lecturer in Communication and Media, University of Liverpool.
• Dr Mark Pogson. Lecturer in Data Science and Communication, University of Liverpool.

The workshop will feature a facilitated activity to develop collaborative grant proposals which apply ABMs to an area of interest chosen by each breakout group.

The day will end with a poster session, which all participants are invited to present at. Posters can be on any topic relevant to social interaction or collective behaviour, not limited to ABMs. A £50 prize is available for the best poster by a PhD student.

The day is fully catered, including lunch, coffee breaks and refreshments during the poster session.

No prior technical experience is required. Theory and simulation are important to the workshop, but we will also look at the role of real-world data and experiments to derive model parameters and validate results. We will also consider example applications like sociology, public health and linguistics.

Background

We identify the micro-scale (e.g. an individual social actor or text) and the macro-scale (e.g. a demographic or shared identity) as being important to many areas of research. In the workshop, we will focus on transitions between the two, termed the meso-scale. For example, how do separate individuals come together to change collective attitudes, and conversely, how do collective attitudes filter back down to separate individuals?

ABMs have the potential to provide a transparent and intuitive predictive framework to explore the meso-scale. ABMs can be integrated with other more common approaches, including textual analysis, ethnography and methods from psychology.

Attending

The workshop will be in person only. Attendance is free but travel expenses are not covered. Please let us know when you register if there are any challenges to you being able to attend (including financial) and we will do what we can to help.

Accessibility information for the venue is available here: https://www.accessable.co.uk/university-of-liverpool/access-guides/school-of-the-arts