Blockmodeling
Also known as: block modeling, blockmodel analysis, generalized blockmodeling, CONCOR
Blockmodeling is a family of methods that simplify a social network by partitioning its actors into positions — groups of actors who are equivalent in their pattern of ties — and summarizing the relations between positions as a compact image, or reduced role structure. Introduced by Harrison White, Scott Boorman, and Ronald Breiger in 1976, it shifts attention from individuals to the structural roles they occupy.
Key highlights
- Reduces large, multi-relational networks to an interpretable role structure, revealing macro-level organization invisible at the tie level.
- Generalized blockmodeling lets the analyst pre-specify and test a hypothesized role structure rather than only discovering one inductively.
- Handles several relations simultaneously, which is essential for role analysis (e.g., authority plus friendship plus communication).
- Provides a bridge between formal role theory and empirical network data.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use blockmodeling when the research question concerns roles and positions — how categories of actors relate to one another — rather than individual centrality, and when you want to reduce a complex network (often multiple relations at once) to an interpretable role structure. It suits role analysis in organizations, kinship, world-system trade, and inter-organizational fields. It is less appropriate when ties are best treated continuously without a positional logic, when the network is too small for stable partitions, or when the substantive interest is in individual-level flow measures such as betweenness. The choice of equivalence type (structural vs. regular) is consequential and must be theory-driven.
Strengths & limitations
- Reduces large, multi-relational networks to an interpretable role structure, revealing macro-level organization invisible at the tie level.
- Generalized blockmodeling lets the analyst pre-specify and test a hypothesized role structure rather than only discovering one inductively.
- Handles several relations simultaneously, which is essential for role analysis (e.g., authority plus friendship plus communication).
- Provides a bridge between formal role theory and empirical network data.
- Results depend heavily on the chosen equivalence definition and the number of positions K, both of which the analyst must justify.
- Exact structural equivalence is rare in real data, so partitions are approximate and the inconsistency they tolerate is a modeling choice.
- Direct optimization is combinatorially hard; heuristic search may find local optima, and solutions can be unstable across runs.
- Classic deterministic blockmodeling provides no probabilistic uncertainty quantification, unlike its stochastic blockmodel counterpart.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
What is the difference between structural and regular equivalence?
Two actors are structurally equivalent if they have ties to exactly the same other actors. They are regularly equivalent if they have ties to actors who are themselves equivalent — capturing the abstract notion of playing the same role even when connected to different specific individuals. Structural equivalence is stricter and rarer; regular equivalence better matches sociological role concepts.
How is blockmodeling related to community detection?
Community detection seeks densely connected, internally cohesive groups (high within-group ties). Blockmodeling seeks positions defined by equivalent patterns of ties, which may be sparse internally — for example a position whose members never connect to each other but all connect to the same superiors. The two answer different structural questions and can yield very different partitions.
What does a stochastic blockmodel add over classic blockmodeling?
The stochastic blockmodel treats block membership as latent and ties as random with block-dependent probabilities, providing model-based estimation, principled selection of the number of blocks, and uncertainty quantification. Classic (deterministic) blockmodeling instead optimizes a fit criterion against ideal block types without an explicit probability model.
Sources
- 1.White, H. C., Boorman, S. A., & Breiger, R. L. (1976). Social structure from multiple networks. I. Blockmodels of roles and positions. American Journal of Sociology, 81(4), 730–780.
- 2.Doreian, P., Batagelj, V., & Ferligoj, A. (2005). Generalized Blockmodeling. Cambridge University Press.ISBN 978-0-521-84085-7
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Blockmodeling. ScholarGate. https://scholargate.app/sociology/blockmodeling