Approaches to data collection and data sources in terrorism research
Thursday June 20, 11.15 - 12.45
Session 1, Auditorium 7, Eilert Sundt building
Chair: Anders Ravik Jupskås
- The Case of Jihadology - Aaron Y. Zelin
- Towards Open and Reproducible Terrorism Studies - Sandy Schumann
- Sources and use of psychological data on radicalisation and terrorism -Helma van den Berg and Dianne van Hemert
- Cause for Pause: Exploiting US-Person Open Source Records to Predict Threat - Darin Challacombe
The Case of Jihadology
Aaron Y. Zelin, Brandeis University
This paper goes to the heart of this conference ‘The data revolution in terrorism research: implications for theory and practice’ by exploring the case of Jihadology, which I founded and have managed the past nine years. It will explore various angles that are worth exploring based off my experiences surrounding issues like why is such a site necessary and/or useful, questions about dissemination and open access, lessons learned about responsibility and interaction with jihadis online, evolution in dealing with having a website that has the largest repository of jihadi content, interactions with governments and technology companies and how they view and deal with my website, as well as conclusions going forward and how it might help other researchers interested in creating primary source-first websites to assist in their research as well as to the benefit of others in the field. Therefore, this paper aims to shed light on this unique case, but also the moral and ethical questions that have come along with maintaining it for almost a decade in a time of changing online environments and more recent calls for censorship.
Towards Open and Reproducible Terrorism Studies
Sandy Schumann, University College London, Isabelle Van der Vegt, University College London, Paul Gill, University College London, and Bart Schuurman, Leiden University
A recent review of articles published between 2007 and 2016 in the most prominent journals in terrorism studies has highlighted that the majority of scholarship relied at least on some kind of primary data; increasingly statistical analyses are used to test hypotheses. We propose that in order for researchers, the research community, and society at large to benefit from the growing prevalence of empirical research in terrorism studies, open science practices—especially such that foster reproducibility—should be implemented systematically. To support this argument, we provide a brief introduction to the open science movement and advances made in other disciplines. Following, we present the results of a survey study (N = 75) that investigated open science activities that terrorism researchers already engage in as well as barriers they see to ‘doing open science’. Considering especially the concerns that respondents voiced in the survey, we introduce practical steps for getting started with open science and describe how researchers can pre-register analyses.
Sources and use of psychological data on radicalisation and terrorism
Dr Helma van den Berg & Dr Dianne van Hemert, TNO, The Netherlands
Despite the increased availability of (empirical) data in the terrorism domain, there is still a paucity of relevant psychological data. These data are essential for building new and validating existing psychological models of radicalisation. We discuss several suggestions for the collection of individual-level psychological data and provide examples of analyses that can be done with these data. Individual-level psychological data that are reviewed include biographical data on known terrorists and radicals, scraped online data from discussion fora, meta-analytic psychological data, survey data from normal populations, and experimental social psychological data from normal populations. We demonstrate how these data can help to strengthen multidisciplinary theory towards more predictive models.
Exploiting US-Person Open Source Records to Predict Threat
Darin J. Challacombe, Fort Hays State University
With the help of social media, rules, and regulars, access to US-based persons’ personal identifying information (PII) has become increasing easy. We used a systematic approach to create our own dataset of US-based sovereign citizens who had committed violent acts during a 10-year time period, and a concurrent dataset of self-proclaimed sovereign citizens who had not committed any act of violence. We applied a structured professional judgment (SPJ) to these sets to post-dict violence (see Challacombe & Lucas, 2018). Through this process, we had two ethical concerns: Are we researchers allowed to exploit this information without concerns? And, at what point do we need to redact or restrict our datasets in order to protect victims or even offenders’ PII? This discussion will outline our data exploitation methods as well as pose questions about ethical concerns with open source data collection.