The complexity of crime network data: a case study of its consequences for crime control and the study of networks.

The field of social network analysis has received increasing attention during the past decades and has been used to tackle a variety of research questions, from prevention of sexually transmitted diseases to humanitarian relief operations. In particular, social network analyses are becoming an important component in studies of criminal networks and in criminal intelligence analysis. At the same time, intelligence analyses and assessments have become a vital component of modern approaches in policing, with policy implications for crime prevention, especially in the fight against organized crime. In this study, we have a unique opportunity to examine one specific Swedish street gang with three different datasets. These datasets are the most common information sources in studies of criminal networks: intelligence, surveillance and co-offending data. We use the data sources to build networks, and compare them by computing distance, centrality, and clustering measures. This study shows the complexity factor by which different data sources about the same object of study have a fundamental impact on the results. The same individuals have different importance ranking depending on the dataset and measure. Consequently, the data source plays a vital role in grasping the complexity of the phenomenon under study. Researchers, policy makers, and practitioners should therefore pay greater attention to the biases affecting the sources of the analysis, and be cautious when drawing conclusions based on intelligence assessments and limited network data. This study contributes to strengthening social network analysis as a reliable tool for understanding and analyzing criminality and criminal networks.