The efficacy of tables versus graphs in disrupting dark networks: An experimental study

An agent tasked with disrupting a dark (covert and illegal) network must first be presented with information about the organization's structure. This paper uses a behavioral experiment to test in a stylized dark network disruption setting whether an agent makes a better disruption choice when the information is displayed in a table or a graph format. We find that subjects make decisions more quickly when presented with graphs, but that subjects perform weakly better with tables. An important lesson is that the display should explicitly distinguish non-existent links from potential links of unknown status.

• A behavioral experiment on dark (covert and illegal) network disruption is conducted.
• Performance under table network displays is compared with that under graph displays.
• Subjects perform faster with graphs but make weakly better disruptions with tables.
• We draw general lessons about dark network disruption.