The Challenge of Studying Violent Radicalization: Potential Barriers and Solutions

We have identified ten major barriers that investigators can find in their way to study violent radicalization: (a) The main goals are blurred; (b) the theory that should be used to support the research is unclear; (c) the access to samples of interest is almost impossible; (d) the similarities between terrorists and other criminal groups are overestimated; (e) choosing the best sample/s for fulfilling the goals is not easy; (f) the nature of field work involves unusually strict ethical protocols; (g) implementing the most appropriate methodology for data collection is complicated; (h) developing the skills needed to talk with violent extremists requires practice and time; (i) the data exploitation strategy must be conscientiously planned; and (j) advances for science and society should be balanced.

This is not a detailed and extensive review of all the barriers and difficulties entailed by the study of violent radicalization. A meticulous analysis would require extra space. Our aim is to give a first step in this direction highlighting the relevance of some of the most serious challenges that an investigator might have to confront and providing possible answers to each of them. In what follows, we will briefly summarize each of these challenges in the form of a general question that needs to be answered and offer some potential solutions (see Figure 1 for a summary).

Figure 1.

Potential Barriers, Questions to be Answered, and Possible Solutions when Studying Violent Radicalization

First: Do I Know What my Goals Are?

The phenomenon of terrorism is not new (see Gómez & Vázquez, Reference Gómez and Vázquez2021 for a brief history), but the studies on this topic increased exponentially after several striking terrorist incidents that took place in the West, as the September 11, 2001 attacks in the United States; the March 11, 2004 attacks in Madrid; or the July 7, 2005 attacks in London (Young & Findley, Reference Young and Findley2011), among others. The subject has been of interest to several disciplines, although it has been mostly addressed by political science and international relations, followed by criminology, penology and law (Haghani et al., Reference Haghani, Kuligowski, Rajabifard and Lentini2022). Researchers have usually worked in disciplinary silos and have been scarcely aware of the relevant contributions from other academic fields (e.g., Silke, Reference Silke and Silke2019).

Articles based on the cross-fertilization between disciplines are extremely rare as well (e.g., Youngman, Reference Youngman2020). Accordingly, we make two suggestions. First, clarify and demarcate the research goals, elaborating objectives that do not overlap with those that can be better achieved by other disciplines and, as a psychologist, prioritize the study of the psycho-social mechanisms underlying violent radicalization. Second, encourage interdisciplinary collaborations.

Second: Do I Need “One” Specific Type of Theory to Support my Investigation?

Understanding why some individuals are willing to self-sacrifice for a group or for their values is a key when studying violent radicalization. Unfortunately, there is not a unique explanatory theory. We offer three possible paths to achieve this end. First, examine existing theories to determine what is valid and can be applied, and what is missing and should be explored. Second, develop new theories. In our case, we did not find previous models considering the synergistic relation that could be established between personal and social identities; and we developed identity fusion theory – a visceral and inviolable feeling of oneness with a group that predicts extreme pro-group behavior – (Gómez et al., Reference Gómez, Chinchilla, Vázquez, López‐Rodríguez, Paredes and Martínez2020; Swann et al., Reference Swann, Jetten, Gómez, Whitehouse and Bastian2012). Third, combine previous theories into new models, like the 3N model did (Kruglanski et al., Reference Kruglanski, Bélanger and Gunaratna2019), which integrates factors related to the need for personal significance, the social networks in which individuals are immersed, and their narratives; or the model of the devoted actor, which applies to individuals fused with a group in which the members share a sacred value – immune to material or non-material trade-offs – (Gómez et al., Reference Gómez, López-Rodríguez, Sheikh, Ginges, Wilson, Waziri, Vázquez, Davis and Atran2017).

Third. What Can I Do if I Don’t Have Access to Samples of Interest?

Perhaps the most insurmountable obstacle is accessing to individuals under risk of radicalization, terrorists, or former terrorists. However, this should not be a full impediment for researchers interested in the study of violent radicalization.

Investigators can conduct paper and pencil and/or online studies with general population. Such studies offer the possibility of using experimental and longitudinal designs to determine causality and are highly useful to overcome the problem of the lack of field data (Arce et al., Reference Arce, Croson and Eckel2011). For instance, the theory, principles, mediators, moderators, and consequences of identity fusion, which has been identified as the top risk factor predicting radical intentions (Wolfowicz et al., Reference Wolfowicz, Litmanovitz, Weisburd and Hasisi2021), were explored and tested in the laboratory years before being finally tested in the field (see Gómez et al., Reference Gómez, Chinchilla, Vázquez, López‐Rodríguez, Paredes and Martínez2020; Gómez & Vázquez, Reference Gómez and Vázquez2015). Other options could be to focus on online radicalization and analyze forums, communities, and web pages where radical content is posted daily (e.g., Gallacher et al., Reference Gallacher, Heerdink and Hewstone2021; Karpova et al., Reference Karpova, Savelev, Vilnin and Kuznetsov2022); or to conduct systematic reviews, meta-analytic research, or replicability studies.

Those researchers that have the opportunity of making fieldwork with samples of interest will have to frame the next barriers.

Fourth: Are All Violent Radicals the Same?

Many governments and international organizations treat terrorists, particularly Islamist terrorists, like members of other criminal groups, such as Latino gangs or delinquent bands (see Gómez, Atran, et al., Reference Gómez, Atran, Chinchilla, Vázquez, López-Rodríguez, Paredes, Martínez, Blanco, Alba, Bautista, Fernández, Pozuelo-Rubio, González-Álvarez, Chiclana, Valladares-Narganes, Alonso, Ruíz-Alvarado, López-Novo and Davis2022). The consequence is an overestimation of the similarities between them. For example, Decker and Pyrooz (Reference Decker and Pyrooz2015) claim that the 100-year tradition studying gangs in the U.S. can inform the study of terrorism. The authors provide a series of lessons to learn from this tradition that are very useful, but they also point out the differences between gangs and terrorist groups. Furthermore, all terrorist groups are not the same (e.g., Bakker, Reference Bakker2015; Weinberg, Reference Weinberg and Silke2019). We can use what we know about certain violent groups to understand others, but we must be careful not to over generalize.

The most obvious way of solving the overestimation problem is to conduct empirical investigations with members of different criminal groups, which allows strong conclusions about the differences and similarities. For example, we recently conducted a study in Spanish prisons, and we found, between other things, that Jihadist terrorists are more fused and willing to self-sacrifice for their most important value – their religion – than Latino-gang members – the honor of the gang, while Latino-gang members are more fused and willing to sacrifice for their most important group – the gang – than Jihadists – Muslims– (Gómez, Atran, et al., Reference Gómez, Atran, Chinchilla, Vázquez, López-Rodríguez, Paredes, Martínez, Blanco, Alba, Bautista, Fernández, Pozuelo-Rubio, González-Álvarez, Chiclana, Valladares-Narganes, Alonso, Ruíz-Alvarado, López-Novo and Davis2022). In another investigation exploring why former terrorists joined radical groups, we discovered that an important percentage of Tamil Tigers – ethno-nationalist, separatist terrorists – joined the group due to environmental pressures, while Jihadists were more likely to enter the group because of reasons related to their identity (Gómez et al., Reference Gómez, Chinchilla, Vázquez, López‐Rodríguez, Paredes and Martínez2020).

When investigators have access to a single group, they can try to replicate the findings obtained by other researchers with other groups to test whether what they have discovered is also valid for the focal group. Finally, when no group is available, researchers can compare already published studies with different samples or use alternative strategies, like social media analysis (e.g., Parekh et al., Reference Parekh, Amarasingam, Dawson and Ruths2018).

Fifth: How Should I Decide the Most Appropriate Sample/s for my Goals?

To choose the right participants, we need to ponder that the nature of the samples restricts the type of conclusions that can be reached. For a comprehensive understanding of the underlying mechanisms related to violent extremism, it is fundamental to study people at different levels of radicalization. For example, individuals under risk of radicalization and terrorists provide highly valuable information on the causal antecedents of violent extremism and the factors that drive individuals to engage in violence, respectively; but none of them have much to say about the mechanisms that sustain deradicalization or disengagement. In contraposition to that, former terrorists can provide rich accounts of the processes motivating disengagement and the peculiarities of life after terrorism, but their explanations of the processes that motivated them to become terrorists might be distorted by lack of memory or by social desirability (e.g., Horgan, Reference Horgan2012). Previous literature acknowledges two further limitations as well: The lack of control groups, which impedes the provision of causal explanations (Freilich et al., Reference Freilich, Chermak and Gruenewald2015); and the lack of empirical studies with women, which limits the conclusions of the studies to men. To select the samples for the research, considering the previous points, we recommend including: (a) Participants at different levels of radicalization; (b) groups of comparison with the same characteristics as the target groups; and (c) women. When none of that is possible, restrict the conclusions of the research accordingly.

Sixth: Which Is the Difficulty to Follow Ethical and Security Protocols?

The study of violent extremism raises serious ethical issues (Taylor & Horgan, Reference Taylor and Horgan2021), and there are no official codes to address them (Conway, Reference Conway2021; Dolnik, Reference Dolnik2000; Morrison et al., Reference Morrison, Silke and Bont2021). Consequently, the ethical planning of the studies tends to be dreadful and prohibitively complex. According to our experience, we suggest a protocol including three basic requirements to avoid undue harm and guarantee knowledge integrity.

First, the risks to the safety and security of participants and researchers must be prevented by adopting strategies tailored to the populations, countries and contexts involved. We must recognize that the participants and researchers are vulnerable in highly specific ways. Physical, psychological, and social risks should be minimized in all the stages of the research. This entails employing interviewers with adequate training to work in high-risk contexts, avoiding techniques that may raise suspicions, providing psychological support during and after the research, or gaining a deep understanding of the socio-political structures of foreign countries and complying with their local legislations, among other things.

Second, participants in violent extremist research have the same rights than those in other types of research, including the right to anonymity, confidentiality, and to be fully informed and provide their voluntary consent for participating. When possible, it is advisable to anonymize the studies by design and to talk with participants in private or semiprivate spaces. Also, the interviewers must ensure that participants are able to consent and comprehend all aspects of the research. In addition, it is important to ensure that researchers’ attitudes or biases do not affect the conduct of the research. And, unless when it is necessary to avoid serious harm, they must maintain absolute confidentiality about the kind of research they are conducting, and the data collected.

Their confidentiality must be guaranteed by official documents.

And third, it is necessary to avert any possible misuse or abuse of the research.

For that, it is extremely useful to keep conversations with representatives of the authorities involved and ensure that they understand the importance of conducting independent analyses and to protect participants and researchers’ safety and rights. Framing the results of the research under the most appropriate light is also necessary. Lastly, counting with the advice of representatives of the rights of the participants is an invaluable asset.

Seventh: Which Is the Best Method to Collect the Data?

The rigor of radicalization research regarding its methodology has been a major concern (Neumann & Kleinmann, Reference Neumann and Kleinmann2013) and has received ongoing criticism (Schuurman, Reference Schuurman2020). Qualitative approaches dominate the field; and an important number of publications rely on secondary sources to support their conclusions. There is also a tendency for authors to work alone rather than collaborate with colleagues – from the same or other disciplines – and to make single contributions instead of conducting continuous programs of research (Schuurman, Reference Schuurman2020). These factors together suggest that a huge number of works on radicalization are methodologically and/or empirically poor. We recommend using triangulation of methods and combining qualitative and quantitative approaches. However, quantitative methods are not easy to implement with samples of participants at different levels of radicalization because of three main reasons: samples are usually small – which reduces the possibility of finding statistical effects; when using big scales, participants might be reluctant to respond to all items – which diminishes their reliability; and some participants might be illiterate – which makes difficult to complete the questionnaires. The potential solutions we offer for each point are: (a) Adding control groups, which increases the statistical power and offers important insights through comparison; (b) using single-item scales or scales including 3–4 items as maximum; and (c) including vignettes or even dynamic measures when possible (e.g., Gómez, Atran, et al., Reference Gómez, Atran, Chinchilla, Vázquez, López-Rodríguez, Paredes, Martínez, Blanco, Alba, Bautista, Fernández, Pozuelo-Rubio, González-Álvarez, Chiclana, Valladares-Narganes, Alonso, Ruíz-Alvarado, López-Novo and Davis2022; Gómez et al., Reference Gómez, López-Rodríguez, Sheikh, Ginges, Wilson, Waziri, Vázquez, Davis and Atran2017; Gómez, Chiclana, et al., Reference Gómez, Chiclana, Chinchilla, Blanco, Alba, Bautista and Pozuelo-Rubio2022).

In addition, when obtaining information from direct conversations with members of the populations of interest, they may give socially desirable answers, lie, or refuse to answer certain questions, which can affect the results and conclusions and compromise the integrity of the research. There are several strategies that can be adopted to prevent these risks, like using lie detection scales, repeating questions with different wording, asking questions indirectly (e.g., third-person wording), or using non-participant observation (e.g., González-Álvarez et al., Reference González-Álvarez, Santos-Hermoso, López-Novo, Chiclana de la Fuente, Pozuelo- Rubio, Fernández-Gómez and Lanza-Comenaux2021).

Eight: Which Are the Appropriated Skills for an Investigator to Talk with Terrorists?

Interviewing individuals under risk of radicalization, former terrorists, and actual terrorists is a big challenge. The lack of experience of interviewers, the reluctance of some investigators to directly interview terrorists, the consequences of the information shared over the interview for the own investigator (negative emotions, anxiety, etc.), or the unwillingness to share the experiences of the interviews with other colleagues are some important limitations for those who directly meet the experience of talking with radicals.

Some authors have identified the type of questions that should be included, some tips for understanding the significance of interviews, how framing initial meetings and reactions, the different interview styles, and also the benefits, challenges and limitations of interviewing terrorists (see Horgan, Reference Horgan2009, Reference Horgan2012). However, we have learned through our experience that the ability to talk to radicals is not merely a list of recommendations for the face-to-face encounter, but that involves more requirements that would facilitate the work of the investigator. In particular, we recommend: (a) Establishing an initial training program for those who will collect the empirical data directly (including simulations); (b) developing periodical meetings between the interviewers during the data collection to share their impressions and concerns; (c) ensuring the implication of all team members in the interviews – either by making interviews directly or by listening to the audios when interviews are recorded; (d) making a final evaluation to identify the factors that facilitated or deteriorated the data collection, which could be useful to improve future research; and (e) sharing the findings with other colleagues, under confidentiality agreements.

Ninth: Which Is the Best Strategy for Optimizing the Data?

When fieldwork is over, it is compulsory to have a plan for obtaining the maximum information from the data and to decide how to exploit it (see for example Schmidt, Reference Schmidt2021). Perhaps, the most important dilemma is to determine what can be published and how, and which information can be provided to governments and institutions without breaking the confidentiality agreement. For the first issue, we strongly encourage researchers to combine field data with violent extremists, exceptionally costly from different points of view, with data from the general population that can be obtained with traditional paper and pencil or online questionnaires (see for example Gómez, Bélanger, et al., Reference Gómez, Bélanger, Chinchilla, Vázquez, Schumpe, Nisa and Chiclana2021). This is a helpful strategy to convince academics and practitioners alike that the findings are valid and relevant from a theoretical and applied perspective. For the second issue, we advise researchers to develop strategies to prevent violent radicalization and promote deradicalization and disengagement, and to share these strategies with government institutions and civil society organizations that aim to combat terrorism using approaches that ensure basic human rights.

Tenth: Is it Worth the Effort?

It is difficult to determine if it is really worth conducting research in this field. From a purely scientific perspective, publishing articles including this type of data is inexplicably difficult in most cases. Nevertheless, researchers must disseminate their findings; and combining our suggestions to the previous points – e.g., having a solid theory, a rigorous method, combining field data with other data, etc. – might facilitate the endeavor ostensibly. From a practical viewpoint, researchers have the opportunity to put their knowledge at the service of developing intervention programs with the potential to greatly benefit society – including individuals at different levels of radicalization. For example, Feddes and Gallucci (Reference Feddes and Gallucci2015) found that between 1990 and 2014, only 12% of evaluations of programs aimed at preventing radicalization and fostering deradicalization included empirical qualitative or quantitative data. Also, a recent meta-analysis from Jugl et al. (Reference Jugl, Lösel, Bender and King2021) encompassing 14,000 reports on the topic of extremism prevention indicated that only nine of them apparently included the evaluation of the results of the programs. This fact highlights the need for a stronger collaboration between the Academy and governmental and civil society organizations.

Understanding the underlying mechanisms leading to violent radicalization is essential to avoid that individuals enter terrorist groups and promote that they abandon violence. Collaboration between researchers and institutions (e.g., governments, law enforcement, and penitentiary institutions) and adopting a multi-disciplinary, -cultural, - theorical, and -methodological approach would provide a richer and more accurate understanding of the underlying mechanisms involved in the dynamic of violent radicalization, contributing to scientific – theoretical knowledge – and social – practical applications – advances. There are two alternatives to fight against violent radicalization: not doing anything or fighting to find a remedy. Fortunately, the number of scientists opting for the second alternative is growing. Here we have combined our field experience with the existing literature to offer a preliminary overview of the main barriers and questions that need to be answered and provide solutions to find this remedy.