Criminal networks are defined by the need of their participants to avoid detection. This need to remain undetected makes any data collection about criminal networks very complicated, yet not impossible. In this chapter, we first review three key areas of research on criminal networks: central actors, networks structures, and network dynamics. We follow with a classification of data sources into text-based, log-based, and cyber-based. Each of these data sources is prone to biases, namely delineation of network dynamics, information not missing at random, spotlight effect, and qualitative context integration. These biases reflect the fact that none of the sources is a primary data source. In the practical part, we demonstrate how to carry out research in the three areas by studying a dataset of each of the three types and showing how the biases therein may be dealt with.
Diviák et al. (Mon,) studied this question.