Conducting evaluations helps us to understand the effects of prevention work aimed at countering radicalization and extremism in the social context. They aim to provide us with answers to frequently posed questions relating to the tangible outcomes of prevention work. Such evaluations, however, are often subject to overblown expectations as to the results they can offer and their feasibility. The justified interest in arriving at solid evidence of their efficacy runs into serious difficulties associated with the planning and implementation of evaluation studies in the area of deradicalization, distancing and radicalization prevention.

One of the aims here is to present and discuss a number of approaches for evaluating the impact of prevention measures in the area of radicalization prevention, deradicalization and demobilization. In this film, we also look into quality criteria and standards for evaluation research. These criteria can help the agencies that contract evaluation studies, as well as their target groups, assess their reliability and significance.

The investigations clearly show that the idea of “evidence-based” prevention only proves viable when evaluation research takes account of the particularities, contradictions and controversies that exist among researchers and practitioners and when these are critically reflected upon. One crucial step here is to clarify contested concepts such as radicalization, prevention (of radicalization) and evidence-basing (for radicalization prevention). In this context, evidence-based means that measures aimed at preventing social problems are developed, implemented and evaluated on the basis of research findings. As prevention can be infinitely expanded in terms of time and target groups, one conceptual and normative question that arises has to do with where (radicalization) prevention should start and finish.

1. A clear assignment. Assignments that have not been clearly defined can result in poor evaluations. The client contracting the evaluation should therefore clearly articulate which type of evaluation is desired when issuing a tender.
2. An evaluation is not an automatism. An obligation to evaluate is desirable. Impact-oriented evaluations are only useful when: first, the target of prevention, the level of prevention, the objective of prevention and the project logic (the effect mechanisms) have been specified; second, the knowledge that the evaluation study aims to acquire has been formulated in agreement with the parties involved; and, third, sufficient financial resources have been allocated for an evaluation that is appropriate for its target.
3. Practice patience. Quick results should neither be expected in prevention work nor in their evaluations. Sponsoring agencies should therefore plan evaluations with a long-term orientation and allot sufficient funding.
4. Consider all relevant data. All of the relevant project data should be useful for an evaluation. In many cases, the most essential data only becomes available once a project has concluded, meaning that accompanying program evaluations should be funded for the period after the measure has ended, as well.
5. A plurality of methods. No patent recipes exist for evaluating radicalization prevention. Choosing the suitable empirical social research methods for evaluations will depend on the concrete findings and on the target of the evaluation.
6. Reflexivity. Policy-makers, society and evaluation researchers should bear in mind that evaluation projects do not simply take a “neutral” and “external” view of practical findings. Rather, they always shape the target they are evaluating to some degree, even in the very selection of the study design. As such, it is necessary for all actors involved to confront and reflect upon this complexity.