Abstrakt
Networks have recently become ubiquitous in many scientific fields. In criminology, social network analysis (SNA) provides a potent tool for analysis of organized crime.
This paper introduces basic network terms and measures as well as advanced models and reviews their application in criminological research. The centrality measures - degree and betweenness - are introduced as means to describe relative importance of actors in the network.
The centrality measures are useful also in determining strategically positioned actors within the network or providing efficient targets for disruption of criminal networks. The cohesion measures, namely density, centralization, and average geodesic distance are described and their relevance is related to the idea of efficiency-security trade-off.
As the last of the basic measures, the attention is paid to subgroup identification algorithms such as cliques, k-plexes, and factions. Subgroups are essential in the discussion on the cell-structure in criminal networks.
The following part of the paper is a brief overview of more sophisticated network models. Models allow for theory testing, distinguishing systematic processes from randomness, and simplification of complex network structures.
Quadratic assignment procedure, blockmodels, exponential random graph models, and stochastic actor-oriented models are covered. Some important research examples include similarities in co-offending, core-periphery structures, closure and brokerage, and network evolution.
Subsequently, the paper reflects the three biggest challenges for application of SNA to criminal settings - data availability, proper formulation of theories and adequate methods application. In conclusion, readers are referred to books and journals combining SNA and criminology as well as to software suitable to carry out SNA.