Social network analysis approaches the social world from a “network perspective” by focusing on the social relations between individual entities or units (such as people, organisations or countries).

Social network analysis encompasses both theory and methods. Methodologically a distinction can be made between more descriptive approaches and more causal (or statistical) approaches. Network analysis can occur at a number of levels, including the dyad (the connectedness between two units), individual (the individual position) and group-level (the network structure as a whole).

Here are some common properties at each of these levels.

Dyadic level connectedness

• Direct tie
• Structural equivalence
• Geodesic distance

Individual network position

• degree, closeness centrality
• betweenness centrality
• structural holes
• diversity

Group level network structure

• Density
• Centralization
• Clustering
• Level of homophily

At each level we might be interested in explaining the emergence of such network properties (antecedents), or focus on their consequences for individual or group outcomes. In my forthcoming chapter in the Oxford Handbook of Social Networks, I try to summarize these different approaches.