By Simon Haux, Svetla Koynova, Lotta Rahlf

This article was first published in German in: Interventionen #19, 2025 by Violence Prevention Network

Is it possible to change the attitudes of individuals who are at the brink of becoming extremists or who have already been radicalised, enabling them to live self-determined lives and contribute constructively to society? What circumstances and conditions would be necessary for this? When an action aims at changing the attitudes, beliefs or motivation of individuals and their families, success and failure are difficult to measure. Yet such actions, as employed in secondary and tertiary prevention, are of great importance for the functioning of a democratic society. That is why the methods for measuring effectiveness in P/CVE are increasingly coming into focus and are the subject of controversial debate. In a politically and socially sensitive and, due to the constraints of project funding in a highly competitive field such as P/CVE, evaluation and quality assurance are increasingly relevant as a means of internal control and external accountability. At the same time, evaluation and quality assurance can provide valuable insights for project development, promote learning processes and thus contribute to strengthening the entire field. Findings on the outcomes of secondary and tertiary prevention are not only relevant for funding bodies and implementing organisations but also for scientific debate.

If they are so useful, why are evaluations of outcomes and impact of counselling work still so rare? The simple answer is that it is difficult to imagine a more challenging subject to study, both methodologically and practically. Most counsellors believe that a simplified, purely quantitative assessment of the ‘success’ or ‘failure’ of pedagogical counselling work is inadequate. Individual case trajectories are often cyclical and different counselling dimensions and goals interact with each other. In addition, the relatively small number of cases within individual counselling centres and projects, combined with the long-term nature of individual disengagement processes, make it difficult to conduct a meaningful outcome assessment within short (often one-year) project or reporting periods. Violent extremists find themselves at the fringes of society and are thus by definition not a large group, and those who agree to counselling are even fewer. Simply establishing a dialogue constitutes a success for the counselling work. ‘Gold standards’ of evaluation research, such as experimental designs with comparison groups, prove to be methodologically challenging with such low n-numbers and ethically problematic when it comes to public safety on the one hand and the right to a meaningful second chance on the other. Comparative studies across projects or even organisations are also hardly feasible due to the stringent data protection standards that must be applied in this field. The contextual differences also limit the significance of any findings. A feasible model for monitoring and measuring the outcomes of counselling work must therefore aim to strike a balance between meaningfulness, reduction of complexity, and resource efficiency for the professionals involved (cf. Klöckner et al. 2021, p. 7f.; Junk et al. 2022).

In order to meet these challenges, there has been increased research in recent years into how counselling work can be structured and documented to generate insights into its effectiveness and impact. One such possibility is offered by the social diagnostic procedure GRIDD PRO – Social Diagnostics, which was developed by Violence Prevention Network gGmbH and supports both the professionalisation of counselling work and the systematic monitoring of its effectiveness. However, this procedure is only partially applicable to the specific challenges associated with assessing family counselling, where external evaluation remains particularly relevant.

GRIDD PRO – Social Diagnostics in counselling

The social diagnostic procedure GRIDD PRO – Social Diagnostics, developed by Violence Prevention Network gGmbH, offers a potential foundation for monitoring and evaluation in secondary and tertiary prevention by structuring counselling activities. GRIDD PRO – Social Diagnostics professionalises the practice of disengagement work by addressing the specific challenges of social work and P/CVE, integrating both in one systematic approach. The resulting structure gives disengagement work a new methodological depth: it enables a holistic understanding of cases, allowing for diagnostically sound, strategic decisions to be made and their effects to be recorded.

The continuous development of hypotheses during the counselling process provides a prerequisite for assessing effectiveness, particularly at the conclusion of a counselling process. This assessment can incorporate both the counsellor’s and the client’s perspective. It can inform monitoring and reporting and provide a basis for evaluations. In addition, these continuous, structured assessments can reveal new mechanisms of action in disengagement and deradicalisation work. This systematic documentation of case work is facilitated by the GRIDD PRO – Social Diagnostics app.

The GRIDD PRO – Social Diagnostics app

GRIDD PRO – Social Diagnostics is based on the ‘application procedure for social diagnostics specifically for the field of disengagement work’ developed by Violence Prevention Network gGbmH (von Berg 2024, p. 27; translated from German). Based on the social diagnostic cycle, a holistic case analysis is first carried out through biographical diagnostics, network, social area, needs, inclusion and resource analysis. This enables the development of hypotheses regarding the outcomes of the planned intervention, which is based on clearly defined goals and success criteria. The process concludes with continuous measurement of effectiveness, in which the instruments used are reapplied to check for changes in the client and to what extent the hypotheses have been realised (ibid.). On the one hand, this measurement of effectiveness can lead to adjustments to the hypotheses and the resulting assistance plan (Violence Prevention Network 2024, p. 9; von Berg et al. 2024, p. 216), but on the other hand, it is also suitable for laying the groundwork for evaluation.

The GRIDD PRO – Social Diagnostics app was developed for the implementation of GRIDD PRO – Social Diagnostics in counselling practice. The app helps counsellors to systematically apply GRIDD PRO – Social Diagnostics in their work. Counsellors choose the relevant sections for each individual case. The ‘Analysis’ section covers analyses of clients’ needs, social inclusion, risks and resources. It encourages the use of specific methods, such as genograms, network maps, timelines, inclusion charts or questionnaires to identify needs. The results of this phase inform the formulation of hypotheses about expected outcomes and a mandate derived from the client’s identified need for assistance. Addition sections include a structured plan of intervention activities, a timetable, as well as intended outcomes and indicators of success (Violence Prevention Network 2024). Counsellors can choose from a variety of predefined approaches or develop new, individual intervention plans.

Assessing effectiveness, enabling evaluation

The evaluation report on federally funded counselling centres (Karliczek et al. 2023, p. 29) proposes a dimensional model of successful disengagement to identify potential goals of disengagement interventions. A ninefold matrix allows 30 development goals with around 100 indicators to be formulated in relation to the (presumed) radicalised individual and their interface with their (social) environment. At the same time, it differentiates between the pragmatic, socio-affective-emotional and ideological-normative levels of disengagement work. The pragmatic level involves working on specific, practical causes that stand in the way of disengagement. The socio-affective-emotional level aims to reduce emotional attachment to the extremist milieu, while the ideological-normative level addresses ideological attitudes (cf. Rabasa et al. 2010, p. 24; Demant et al. 2008). This type of structure also allows statements on the effectiveness of disengagement work to be focused thematically and the different levels of goals to be analysed in a targeted manner.

GRIDD PRO – Social Diagnostics illustrates how a participatory process with counsellors can lead to a structured formulation of goals, thereby supporting a nuanced and practice-oriented design of the counselling process. First, overarching outcome goals are identified in areas in which positive change is sought for individual clients, depending on the diagnostic conclusions. These outcomes are subdivided into 15 tangible, action-oriented intervention goals. These goals are in turn broken down into 170 success criteria (indicators) in order to measure the clients progress towards the defined goals. While some indicators can be straightforwardly assessed (e.g. participation in counselling sessions), others are based on the counsellors’ qualitative assessment and professional experience (e.g. personal openness or biographical storytelling). The combination of outcomes, intervention goals and indicators of success forms the basis for the systematic monitoring of effectiveness, capturing both objective and subjective components. Such an approach helps to uncover individual counselling processes’ impacts without obscuring their complexity.

Karliczek et al. (2023, pp. 105–106) describe a structured assessment form that can be used to record the initial situation at the beginning of the intervention and the state achieved afterwards for each development goal. They emphasise that this also allows for long-term developments to be traced over time, thus enabling impact-oriented evaluation. However, this application was not possible within the time frame of their study; the potential is discussed primarily in conceptual terms (ibid.). GRIDD PRO – Social Diagnostics offers an alternative for monitoring and evaluation, which has not yet been tested in the realm of civil society counselling. For this, it is crucial that monitoring with GRIDD PRO – Social Diagnostics is carried out regularly. Von Berg et al. (2024, p. 217) point out that it is necessary to repeat the social diagnostic cycle for evaluations, as a one-off measurement of effectiveness only provides a snapshot and cannot capture long-term effects. Continuous monitoring also allows counsellors to review their assumptions, update diagnoses and improve measures.

It would be particularly insightful to compare the possibilities of outcome evaluation based on data from GRIDD PRO – Social Diagnostics with those of the structured assessment form developed by Karliczek et al. Both approaches present promising foundations for individual case analysis and impact assessment, while at the same time tapping into the potential for case-independent abstraction in monitoring and evaluation systems. GRIDD PRO – Social Diagnostics offers the advantage of a structure that has already been introduced and implemented in practice, enabling precise operationalisation and supporting efficient implementation and monitoring through the accompanying app. The dimensional model developed by Karliczek et al. can serve as a supplement to review and further develop processes.

In conclusion, outcome assessment is an integral part of the social diagnostics cycle.
Continuous, long-term data collection is a prerequisite for the meaningful assessment of effectiveness. To this end, it is essential that counsellors document and reflect on their practice using the GRIDD PRO – Social Diagnostics app (Violence Prevention Network 2024). A structured framework based on the GRIDD PRO – Social Diagnostics approach can facilitate the subsequent analysis of this data for monitoring and evaluation. This is the key not only to measuring outcomes and impact, but also to the success of social diagnostics as a method.

Measuring effectiveness in counselling work with relatives

As Karliczek et al. (2023, p. 30) note, counselling relatives accounts for a significant part of the work done by counselling centres in the field of P/CVE. Working with relatives has different objectives than working with index persons themselves, which is why monitoring must be different and the recording systems described above cannot be transferred without adaptation. For this reason, external survey methods will probably continue to play a greater role in the future. Nevertheless, individual topics and categories of social diagnostics are of course also suitable for working with relatives, especially those relating to relationship dynamics. Nevertheless, it should be noted that the intensity of the counselling relationship in family counselling is often not comparable to the relationship between index persons and counsellors: contacts are often less frequent, sometimes even sporadic. Family counselling also does not have the potential to have a comparably strong effect on index persons. Here, the assumptions about effectiveness differ considerably. 

Diagnostic procedures focused on index persons are therefore generally unsuitable for assessing the effectiveness of these services. Measuring the effectiveness of counselling for relatives of extremists is also particularly challenging because it takes place in a highly complex and often emotionally charged environment. Here, too, as in the case of counselling for index persons, counselling goals are often process-oriented and vary from person to person. They range from short-term stabilisation of the person seeking advice to promoting confidence in their actions and supporting the distancing processes of index persons. In order to deal with the process-oriented and unpredictable nature of enquiries, evaluators workiing in more innocuous areas such as care counselling (Wolff et al. 2022) sometimes ask directly about the satisfaction of those receiving counselling.

However, in the context of P/CVE, relatives find themselves in exceptional situations, characterised not only by worry and feelings of guilt, but also by conflicts of loyalty. All of this makes standardised evaluation difficult. A satisfaction survey, which evaluators would normally aim to conduct immediately after an interview, may seem absurd to those seeking advice in such a context and can significantly undermine their trust in the counsellor’s empathy. If the evaluators were to attempt to organise a classic follow-up, the risk of an immediate loss of trust would be reduced, but the willingness to participate would probably still be very low in a process characterised by such ambivalence. In addition, even less information about the client’s trajectory may be available in classic family counselling. Moreover, even more limited information is available on long-term developments or feedback on the effects on the radicalised person themselves, due to data protection and security reasons. This is frequently also the case in situations where the index persons are accompanied by the same organisation in their distancing process. Measuring effectiveness therefore requires not only a sensitive methodological approach, but also a deep understanding of the psychosocial dynamics in family contexts of radicalised environments.

So what can be done from an evaluative perspective when the focus is deliberately not on the radicalised person themselves, but on their family environment? The psychological stresses mentioned above, characterised by uncertainty, shame and ambivalent feelings, make evaluation difficult, but the satisfaction of relatives with the counselling process can be an important indicator of effectiveness nonetheless. Methodologically, a multi-stage approach is recommended here: low-threshold, voluntary and anonymous feedback forms after the counselling session has been completed can provide initial indications, supplemented by qualitative interviews or focus groups with former clients after an appropriate period of time, e.g. one year, provided they are willing to participate. Care should be taken to design the questions sensitively in order to avoid attributing blame and undermining trust.

However, a narrative approach, which can be integrated into the counselling methodology towards the end of the family counselling session itself, also offers good chances of success. In this approach, family members are informed that the data will be used anonymously for quality assurance and evaluation purposes. In this narrative approach, relatives can describe their journey in their own words. This allows valuable information on perceived effects to be collected without additional effort or loss of trust, while at the same time providing a framework that does justice to the complexity of the case when recording it. Although standardisation cannot be achieved, the data collected provides information on the actual effects, which can vary greatly from case to case, e.g. the emotional stabilisation, the strengthening of relatives’ long-term resilience or the immediate psychological relief provided by counselling. If the relatives are motivated, a similar (albeit reduced) goal achievement and success factor grid, such as GRIDD PRO – Social Diagnostics or the dimension model, can also be used. The social environment can also report changes in its relationship with the radicalised person using a genogram, so that these changes can be evaluated.

Measuring impact in the context of referral counselling

Measuring outcomes is particularly complex in the case of referral counselling. In such processes, contact with those seeking advice is usually very brief and thematically very limited, as the primary goal is to clarify the specific concern, briefly outline the contextual conditions and, finally, to refer the person to a suitable support service (in terms of content or region). In the context of P/CVE, it is often primarily relatives who turn to the referral services. As discussed, the threshold for making contact is often already high: Many relatives are ashamed, fear escalation or have doubts about their own assessment of the situation. The initial contact is thus brief and selective, which makes in-depth evaluation even more difficult. Since the actual counselling only begins at the local counselling centres, there are also data protection challenges regarding the flow of information and the possible tracking of effects beyond the counselling process.

In terms of their logic of effectiveness, referral consultations have much more in common with so-called clearing units. Here, the ‘success’ of case clearing is measured by simply quantifying the number of cases closed, whereby both the start of a consultation process or referral to another competent authority and the determination that no further consultation is required can be considered a successful conclusion of the clearing process. Only the inability to reach an assessment of the facts would therefore be considered a ‘failure’. Accordingly, the quantitative success rate is usually 100 per cent.

For a meaningful evaluation of pure referral counselling, an additional qualitative survey that is not results-oriented is therefore necessary. Information about meeting the needs of those seeking advice (usually relatives) and procedural issues, such as the time span until successful referral, the change or presumed expansion of access channels, and the geographical distribution of needs, are particularly important. These can usually be collected indirectly via the cooperating referral counselling structures, which is why the broad involvement of the counselling partners involved is essential for a good evaluation. However, the diversity of perspectives and actors involved necessitates that evaluations of the referral process itself be planned for comprehensive survey periods.

Conclusion

Measuring the effectiveness of secondary and tertiary prevention is methodologically extremely challenging given the complex individual trajectories, the small case numbers and the high data protection requirements. However, approaches such as GRIDD PRO – Social Diagnostics or the Dimensions Model are increasingly being used to develop practical tools that systematically measure effectiveness while strengthening the professionalism of counselling work. Continuous documentation and reflection by professionals are proving to be the key to meaningful evaluations. For areas such as family or referral counselling, adapted, sensitive methods are needed that do justice to the specific dynamics. However, the approaches discussed in this article should also be seen as the product of a successful and close intertwining of research and practice. In order to further develop these approaches in a sustainable manner and to fully exploit their potential for practice, however, long-term and reliable funding for research projects that consistently advance effectivness measurement in this highly relevant field of action is essential.

Authors:

Svetla Koynova heads the Research Department at Violence Prevention Network gGmbH together with Maximilian Campos Ruf and is responsible for the strategic orientation of research initiatives in this context. She studied political science with a specialisation in sociology of religion and intercultural and transcultural management (University of Freiburg and Sciences Po Aix) and has extensive experience in monitoring, evaluation and learning processes as well as in the study of group-focused enmity, conspiracy beliefs and radicalisation. In her role at Violence Prevention Network, she contributes significantly to the further development and scientific foundation of practical work processes, particularly in secondary and tertiary extremism prevention.

Simon Haux coordinates secondary and tertiary prevention projects at Violence Prevention Network gGmbH, particularly in the areas of religiously motivated extremism, conspiracy theories and online radicalisation. After studying political science in Berlin and Buenos Aires, he previously worked on violence prevention approaches in German development cooperation and the monitoring of bilateral projects to support vulnerable groups in the Western Balkans.

Lotta Rahlf is pursuing her doctorate at the Leibniz Institute PRIF – Peace Research Institute in Frankfurt am Main as part of the Marie Skłodowska-Curie Actions (MSCA) doctoral network VORTEX, which is funded by the European Union’s Horizon Europe Research and Innovation Programme. She is investigating the institutionalisation of evaluation and quality management in the field of P/CVE in European comparison. Previously, she worked on the PrEval project, a research and transfer project that develops and refines formats and structures for strengthening evaluation and quality assurance in extremism prevention, democracy promotion and civic education in Germany.

Literature

Demant, Froukje, Marieke Slootman, Frank Buijs, and Jean Tillie. 2008. Decline and Disengagement: An Analysis of Processes of Deradicalisation. Amsterdam: IMES Report Series.

Junk, Julian, Svetla Koynova, Vivienne Ohlenforst, Maximilian Ruf, and Lea Scheu. 2022. Evaluation and Quality Assurance in the Cooperation between Security Authorities and Civil Society Actors in Secondary and Tertiary Extremism Prevention. PRIF Report 8/2022. Frankfurt am Main: Leibniz Institute Hessian Foundation for Peace and Conflict Research (HSFK).

Karliczek, K. M., Ohlenforst, V., Schaffranke, D. & Walkenhorst, D. (2023). Evaluation bundesfinanzierter Beratungsstellen, Abschlussbericht der Evaluation der Beratungsstellen zur Distanzierung und Deradikalisierung vom islamistischen Extremismus (Beiträge zu Migration und Integration, Band 12). Nürnberg. Bundesamt für Migration und Flüchtlinge.
https://doi.org/10.48570/bamf.fz.beitr.b12.d.2023.beratungsstellen.1.0

Klöckner, Mona, Svetla Koynova, Johanna Liebich, and Lisa Neef. 2021. Experiences from the evaluation planning of an exit programme: Prerequisites for measuring effectiveness in tertiary extremism prevention. PRIF Report 6/2021. Frankfurt am Main: Leibniz Institute Hessian Foundation for Peace and Conflict Research (HSFK). https://doi.org/10.48809/prifrep2106

Rabasa, Angel, Stacie L. Pettyjohn, Jeremy J. Ghez, and Christopher Boucek. 2010. Deradicalising Islamist Extremists. Santa Monica, CA: RAND Corporation.

Violence Prevention Network. 2024. GRIDD PRO – Social Diagnostics: Professional Support for Individual Empowerment, Integration, Distancing and Development: Resource-Oriented Tool for Extremism Prevention. https://violence-prevention-network.de/angebote/projektuebersicht/gridd-pro-soziale-diagnostik/?lang=en (last seen 31 October 2025)

Von Berg, Annika, Dennis Walkenhorst, Gloriett Kargl, and Maximilian Ruf. 2023. Soziale Diagnostik in der Extremismusprävention – Diagnose, Fallverstehen, Intervention und Wirkungsmessung. Ideologie und Gewalt – Schriften zur Deradikalisierung. Springer Fachmedien Wiesbaden. https://doi.org/10.1007/978-3-658-42427-5.

Wolff, Julia Katharina/ Pflug, Claudia (2022). Pflegeberatungsstrukturen für pflegebedürftige Menschen mit speziellen Versorgungsbedarfen. In: Jacobs, K., Kuhlmey, A., Greß, S., Klauber, J., Schwinger, A. (eds.) Pflege-Report 2022. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-65204-6_16