This project builds off previously-funded NIJ datasets (Profiles of Individual Radicalization in the United States (PIRUS) and the Social Networks of American Radicals (SoNAR)) to create a relational database that will include event-level information on approximately 700-1,000 U.S.-based extremist plots from the years 1990-2021. These event data will be linked to data on the individuals and social networks contained in the PIRUS and SoNAR datasets, respectively, in order to build a fully relational database of radicalization characteristics, social-network dynamics, and event-level details and outcomes.
The purpose of this new relational database will be to better understand the differences between radicalization and mobilization indicators, as well as violent extremist offenders and nonviolent extremist offenders. Data collection will proceed in multiple waves in which teams of trained researchers will use open-sources, including media reports, unsealed court documents, unclassified government reports, and other open-source archived content, to gather and organize quantitative data based on an event-level codebook of approximately 75 variables. To strengthen the reliability and validity of the analyses, at least 25% of the data will be double coded, and the project team will dedicate the last phase of data collection to conduct rigorous quality control. To analyze these data, the project team will use three methods: descriptive and bivariate analyses; advanced regression methods with regularization; and multistep configurational methods on a subset of 50 event cases. The research team will leverage START’s training development capabilities to create an online training series. This series will aim to educate law enforcement and criminal justice professionals of the risk factors associated with domestic radicalization to violent extremism and identifying strategies for effective prevention and intervention.