Thursday June 20, 13.45 - 14.15
Session 2, Auditorium 7, Eilert Sundt building
Chair: Elena Savoia
- Using an Experimental Design to Test Terrorism Risk Assessment Tools - Sanaz Zolghadriha, Paul Gill and Noemie Bouhana
- Risk terrain modelling of violent dissident Republican activity - Zoe Marchment, Paul Gill and John Morrison
- Application of terrorism risk assessment tools to lone-actor data - Nadine Salman
- The GRIEVANCE dictionary: understanding terrorist language use - Isabelle van der Vegt, Bennett Kleinberg
Using an Experimental Design to Test Terrorism Risk Assessment Tools
Sanaz Zolghadriha, Paul Gill and Noemie Bouhana, University College London
The need to increase our understanding of the risk assessment and threat management of terrorists is vital to the reduction of extremist violence against non-combatants. Currently, a number of risk assessment tools are developed, such as the Terrorist Risk Assessment Protocol 18, the Identifying Vulnerable People tool, the Extremism Risk Guidelines, and the Violent Extremist Risk Assessment guide. However, there is a lack in the literature to systematically assess the reliability and validity of these tools using scientific rigorous research. Using a range of cases that include true positives (actual violent extremists) and true negatives, we will test the reliability and validity of commonly used risk assessment tools. Different measures will be applied to assess the inter-rater reliability of these tools, (e.g. percentage agreement, Kappa, ICC), in addition to testing which tools consistently score highly on the inter-rater reliability measures. The use of false positives and counter-factual cases will allow for tests of validity. The equity of the tools (i.e. how applicable the tools are to different categories of violent extremists) will be tested through the inclusion of cases that range across different ideologies and terrorist type.
Risk terrain modelling of violent dissident Republican activity
Zoe Marchment, Paul Gill, University College London, and John Morrison, Royal Holloway University
There is extensive research to suggest that most terrorist offenders are rational and purposeful in their decision making. They will make carefully calculated decisions that are likely to increase their probability of success, based on perceived rewards, effort and risk. There have been several analyses demonstrating spatial and temporal variation in risk of terrorist attacks, and most conclude that terrorism is spatially concentrated. However, these spatial analyses were unable to identify any potential correlates of these hotspots – just the fact they exist.
Fortunately, in the study of urban crime, risk terrain modelling (RTM) was created to assess risk by analysing the level of opportunity a location may offer to an offender seeking a target. Each location has an associated value to an offender, which is determined by the opportunity for crime that it offers. RTM can be used to identify the locations that have the greatest perceived opportunity and therefore pose the highest level of risk.
This study uses RTM to examine the influences of social and physical context on target selection for Violent Dissident Republican activity in Belfast. This method identified multiple significant risk factors for bombings and bomb hoaxes, and differences between the two incident types. The model shows good predictive accuracy in identifying the areas of a city most at risk of a terrorist attack and could be a useful tool in guiding targeted responses to associated threats.
Application of terrorism risk assessment tools to lone-actor data
Nadine Salman, University College London
Currently, a key priority of counter-terrorism strategies is the identification of high-risk individuals, to facilitate intervention before they can carry out attacks. To this end, several structured professional judgment tools have been developed to help assessors estimate individuals’ risk, comprising indicators based on data and evidence from known terrorists. While individual terrorism risk assessment tools have been evaluated, there has not yet been a comparative evaluation of these tools. This study compares and evaluates different tools, including the TRAP-18, ERG22+/VAF, and the VERA-2R, using data from an open-source sample of US and European lone-actor terrorists. This analysis compares the prevalence of each tool’s indicators in the sample, as well as any differences in their prevalence depending on the extremist ideology (jihadist, far-right or single-issue). The aim of this research is to assess the validity of the different terrorism risk assessment tools, and examine the suitability of each tool in the context of different ideologies.
The GRIEVANCE dictionary: understanding terrorist language use
Isabelle van der Vegt and Bennett Kleinberg, University College London
With the rise of the internet, terrorists have established a significant presence online. Much of the material posted online by terrorist and extremist entities is linguistic in nature, such as forum-posts, tweets, and manifestos. It has arguably become increasingly important to understand radical, extremist and terrorist language use. A common procedure for doing so is to measure linguistic and psychological variables in texts with a dictionary. However, some limitations persist when standard or custom dictionaries are used for this purpose. First, standard psycholinguistic dictionaries (such as Linguistic Inquiry and Word Count) are not developed for assessing terrorist language and therefore do not measure constructs that may be of interest (e.g. aggression, frustration). Second, it is often unclear how and why words are selected for custom dictionaries that are specifically developed for linguistic analysis of terrorist texts. This project addresses these limitations by developing a dictionary which can be applied to understand texts written by extremist, terrorist and other aggrieved individuals. The dictionary will be openly available, and the process of its development is transparently documented. By doing so, we aim to do our part in making sense of the increasingly available (linguistic) data in the field of terrorism studies.