Visualizing the relationship among indicators for lone actor terrorist attacks: Multidimensional scaling and the TRAP-18

Abstract

This validation study analyses data from a sample of North American terrorist attackers (n = 33) and non-attackers (n = 23) through the lens of the Terrorist Radicalization Assessment Protocol (TRAP-18; Meloy, 2017) utilizing a multivariate statistical approach – multidimensional scaling – to visualize potential clustering (co-occurrence) of risk factors. Rarely done in terrorism research, the results plotted in two-dimensional space show the clustering and co-occurrence of most of the eight proximal warning behaviors among the attackers, but not among the non-attackers, and less of a clustering and association of distal characteristics, but their presence in both attackers and non-attackers. These findings provide further empirical support for the rational-theoretical model of the TRAP-18, a structured professional judgment instrument for threat assessment of lone actor terrorists. It advances the quantitative analysis of operationally relevant and behaviorally observable indicators for use by law enforcement and counterterrorism professionals and their consultants. Findings are discussed in relation to other research on pre-offense behaviors of lone actor terrorists, and recommendations are made for both operational use and further research.