‘Between theory and evidence: causal analysis in criminological research’

Criminology faces the task of precisely describing causal relationships between individual, social and institutional factors. Whether in the evaluation of laws and crime prevention measures, the investigation of social determinants of crime or in intervention research, robust causal analysis forms the foundation for evidence-based practice and theory formation.

However, rigorous experimental designs in criminological research are often ethically or practically unfeasible, while non-experimental data often provide only a static ‘snapshot’ of dynamic processes. In addition, there are fundamental methodological challenges in observational studies, such as (unmeasured) confounding, collider and selection bias, measurement error, reverse causality, and problems of generalisability.

Accordingly, both the theoretical foundation of causal assumptions (e.g., through DAGs or the Potential Outcomes Framework) and research designs for identifying causal effects (such as RCTs, quasi-experimental approaches, panel designs or survey experiments) and statistical methods (matching, instrumental variables, structural equation models, data-driven analyses and machine learning) are increasingly coming to the fore.

The aim of the conference is to open a dialogue on standards, innovations and limitations of causal analytical research in criminology. We invite you to join us at our premises in Hanover from 8 December 2025 (starting at approx. 10 a.m.) to 9 December 2025 (ending at approx. 3 p.m.). Interested speakers from various disciplines are cordially invited to present their work on theories, methods and applications of criminological causal analysis. Abstracts of no more than 200 words can be submitted until 19 October 2025 at MethodLab@kfn.de.

If you are interested in attending without presenting a paper, please also contact us by email. However, due to limited space, we will only be able to send out confirmations/rejections after reviewing all submissions.

‘Between theory and evidence: causal analysis in criminological research’

Criminology faces the task of precisely describing causal relationships between individual, social and institutional factors. Whether in the evaluation of laws and crime prevention measures, the investigation of social determinants of crime or in intervention research, robust causal analysis forms the foundation for evidence-based practice and theory formation.

However, rigorous experimental designs in criminological research are often ethically or practically unfeasible, while non-experimental data often provide only a static ‘snapshot’ of dynamic processes. In addition, there are fundamental methodological challenges in observational studies, such as (unmeasured) confounding, collider and selection bias, measurement error, reverse causality, and problems of generalisability.

Accordingly, both the theoretical foundation of causal assumptions (e.g., through DAGs or the Potential Outcomes Framework) and research designs for identifying causal effects (such as RCTs, quasi-experimental approaches, panel designs or survey experiments) and statistical methods (matching, instrumental variables, structural equation models, data-driven analyses and machine learning) are increasingly coming to the fore.

The aim of the conference is to open a dialogue on standards, innovations and limitations of causal analytical research in criminology. We invite you to join us at our premises in Hanover from 8 December 2025 (starting at approx. 10 a.m.) to 9 December 2025 (ending at approx. 3 p.m.). Interested speakers from various disciplines are cordially invited to present their work on theories, methods and applications of criminological causal analysis. Abstracts of no more than 200 words can be submitted until 19 October 2025 at MethodLab@kfn.de.

If you are interested in attending without presenting a paper, please also contact us by email. However, due to limited space, we will only be able to send out confirmations/rejections after reviewing all submissions.