Analyze Brokerage and Structural Holes
Identify actors positioned between otherwise separated contacts or groups, compare brokerage measures, and interpret opportunity without turning network position into a claim about behavior or value.
Method sources
By the end of this tutorial
- 1Distinguish brokerage opportunities from betweenness centrality and ordinary high degree.
- 2Use structural-hole and group-aware measures that match a stated relational process.
- 3Interpret broker positions with sensitivity checks, context, and a responsible evidence boundary.
SNA
Network specification
Analysis scenario
A university maps instructional-advice seeking among 36 educators in science, language, and student-support departments. Leaders want to know which educators could connect knowledge across departments, while avoiding the assumption that a structural opportunity proves that a person actually mediates information.
Nodes
The 36 educators employed in the three participating departments on the census date; department membership is recorded as a node attribute.
Ties
A directed binary tie from educator A to educator B means A reported seeking instructional advice from B during the previous eight weeks.
Network type
One-mode, directed, binary whole network with a fixed three-department boundary, one recall period, department attributes, and nonresponse recorded separately from no tie.
Step-by-step tutorial
Define the brokerage opportunity
State what must flow, who could mediate it, and whether direction matters. Enumerate directed two-paths from a source through a potential broker to a recipient, then flag cases where source and recipient lack the direct tie relevant to your theory.
Checkpoint
Every counted opportunity has an eligible source, broker, and recipient, and its direction and missing-data rule match the advice-seeking question.
Compare complementary measures
Calculate degree and betweenness beside an ego-network measure such as constraint or effective size. Treat low constraint as access to nonredundant contacts, not as a synonym for all shortest-path brokerage, and inspect the actual local configurations behind each score.
Checkpoint
You can explain why a high-degree actor, a high-betweenness actor, and a low-constraint actor may be three different people in the same network.
Classify group-aware roles
Use the pre-specified department attribute to classify mediation opportunities with a framework such as Gould-Fernandez brokerage. Separate within-group coordination from roles that receive from, send to, or connect across groups, and report the eligible opportunity count for each role.
Checkpoint
Every role label follows the source, broker, and recipient group memberships, and departments were defined before viewing brokerage results.
Check robustness and add context
Repeat the analysis under plausible tie thresholds, missing-data treatments, and direction choices. Compare observed role counts with suitable reference networks, then use interviews or process records to learn whether structural opportunities were recognized, used, shared, or burdensome.
Checkpoint
The conclusion identifies which broker rankings survive sensitivity checks and keeps structural opportunity separate from demonstrated mediation, performance, or influence.
Interpret with care
A broker score describes position under a particular boundary, tie definition, and group classification. High betweenness can reflect control over shortest paths, while low constraint emphasizes nonredundant ego contacts; neither alone shows that information moved through the actor.
Brokerage can provide access and coordination opportunities, but it can also create workload, dependence, conflict, or fragility. Group-aware roles are meaningful only when group categories are theoretically relevant and not constructed after seeing who appears central.