Network & Interaction Modelling
Analysing relationships, flows and dependencies between connected parts.
At a glance
- Produces
- Descriptions of structure, dependency, and the routes along which change would propagate.
- Depends on
- An agreed meaning for a connection, and explicit treatment of the links that are missing.
- Fails when
- An edge is analysed before anyone has established what it represents.
Network and interaction modelling treats a system as a set of entities and the relationships between them, then examines how that structure shapes diffusion, influence, coordination and resilience. It is particularly useful where outcomes depend on who is connected to whom.
How we approach it
Networks are usually drawn before they are measured. We try to establish what a connection means, and how it was observed, before analysing it.
- The meaning of an edge is defined before analysis. Influence, information flow, contractual dependency and physical connection are different networks even when they share the same nodes.
- Structure and dynamics are considered together. Static snapshots are useful, but most questions of interest concern how structures change or how flows move through them.
- Uncertain and incomplete network data is treated explicitly, since missing edges change conclusions about centrality and vulnerability.
- Findings are tested against non-network explanations before being reported as structural.