Network analysis, explained
Every term you meet when you start reading a network, answered in its first paragraph and then worked through on a real map. The examples are the classic datasets of network science (Krackhardt's kite, Zachary's karate club, the Florentine families), computed live by the same code that runs Netgraf's insights.
- What is a sociogram?A sociogram is a diagram of the relationships in a group: people as points, ties as lines. What it shows, how to read one, and how to make one free.
- What is social network analysis?Social network analysis studies the ties between people rather than the people alone. The core ideas, the key measures, and three classic examples you can open.
- What is a relationship map?A relationship map is a diagram of how people or organisations are connected. What goes on one, when to use it, and examples from business, history and TV.
- Degree centralityDegree centrality is the number of ties a person has. The formula, the normalised version, in-degree and out-degree, and a worked example on Krackhardt's kite.
- Betweenness centralityBetweenness centrality measures how often a person sits on the shortest path between two others. The formula, why the Medici top it, and how to find brokers.
- Closeness centralityCloseness centrality measures how few steps a person needs to reach everyone else. Classic and harmonic closeness, why they differ, and a worked example.
- Eigenvector centrality and PageRankEigenvector centrality scores a person by how well connected their contacts are. How it relates to PageRank, why PageRank is safer on real maps, and an example.
- Network densityNetwork density is the share of possible ties that actually exist. The formula for directed and undirected networks, what counts as dense, and an example.
- Communities and modularityA community is a group of people tied more to each other than to the rest. How modularity and the Louvain method find them, shown on Zachary's karate club.
- Brokers, bridges and structural holesA broker links groups that would otherwise be apart. Structural holes, weak ties, and cut points (the people whose removal splits a network), with an example.
- Clustering coefficientThe clustering coefficient measures how many of a person's contacts know each other. Local and global clustering, the formula, and a worked example.
- Network centralizationCentralization measures how much a network is organised around one person. Freeman's degree centralization, how to read it, and a worked example.
- AssortativityAssortativity measures whether well-connected people tie to each other or to the less connected. Newman's coefficient, how to read it, and an example.
- Ego networksAn ego network is one person, their direct contacts, and the ties among those contacts. How to collect and map one, and what it reveals.
Draw your own
Free, no account, and private by default. Start blank, import a spreadsheet, or describe the network in words.