Computational Sociology

Computational sociology uses computational methods — agent-based modelling, network analysis, simulation, and large-scale digital data — to study social processes and emergence.

Scope

It covers agent-based models of social dynamics, social network analysis, the analysis of 'big' behavioural data, and the simulation of how micro-interactions produce macro-patterns.

Core questions

• How do macro social patterns emerge from micro interactions?
• How can social processes be simulated?
• What can large-scale digital data reveal about society?
• How do networks shape social outcomes?
• How can computation complement sociological theory?

Key concepts

• Agent-based modelling
• Emergence
• Social network analysis
• Big data
• Simulation
• Micro-macro link

Key theories

Agent-based modelling

Macy and Willer showed how agent-based models let sociologists study how individual actions generate emergent social structure.

Computational social science

Lazer and colleagues argued that large-scale digital data and computation open a new, data-rich social science.

History

Growing from social simulation and network analysis, computational sociology was crystallized by Macy and Willer's 'from factors to actors' (2002) and the computational-social-science agenda (Lazer et al. 2009), now expanding rapidly with digital trace data and machine learning.

Debates

Can big data replace theory?

Whether data-driven computational methods complement or threaten to displace theory-driven sociological explanation.

Key figures

• Michael Macy
• Robb Willer
• David Lazer

Seminal works

• macy-willer-2002
• lazer-2009

Frequently asked questions

What is agent-based modelling?

A simulation method in which many interacting 'agents' follow simple rules, allowing researchers to study how macro-level patterns emerge from micro-level behaviour.