Research
My research spans Bayesian methodology, research software engineering, and applied data science across scientific domains. During my doctoral work I developed a simulation-based method for expert prior elicitation, translating domain knowledge into prior distributions for Bayesian models, and built elicito, an open-source Python package published on PyPI and conda-forge, with documentation, CI testing, and development environment in GitHub.
Beyond pure method development, I have worked in different domains. In climate science I developed data pipelines to assimilate ground-based and satellite observations of greenhouse gas concentrations into data products for earth system modelling groups. In social psychology I contributed to the full research cycle (experimental design, data collection, statistical analysis, and reproducible implementation) and developed mathematical and computational models to formalize verbal theories of truth judgments.
Doctoral thesis
Bockting, F. (2026). Simulation-based expert prior elicitation: Method and software development. Doctoral Thesis. TU Dortmund University.
PDF
Articles
Bockting, F. & Bürkner, P. C. (2025). elicito: A Python package for expert prior elicitation. arXiv preprint.
ArXiv
Talks & Presentations
Bockting, F. (2025). Contributed talk: Predictive prior elicitation — state of the art and current challenges. Workshop on Bayesian Modelling, Lund University.
Slides
Bockting, F., Radev, S. T. & Bürkner, P. C. (2024). Contributed talk: Normalizing flows for simulation-based expert prior elicitation. MathPsych — Society for Mathematical Psychology.
Slides
Bockting, F., Radev, S. T. & Bürkner, P. C. (2024). Contributed talk: Simulation-based prior knowledge elicitation for parametric Bayesian models. ISBA — International Society for Bayesian Statistics.
Slides
Bockting, F. & Heck, D. W. (2021). Measuring individual differences in the truth effect: A formal analysis. Fast Talk at MathPsych.
Book chapter
Stephan, A., Walter, S., Anton, T., …, Bockting, F., …, & Schütze, P. (2021). Nachwort. In Turing, A. M. Computing Machinery and Intelligence. Können Maschinen Denken? (pp. 131–201). Reclam.
Publisher
Courses
Introduction to Python · 2023/24, 2024/25 · TU Dortmund University
An undergraduate and graduate course covering Python fundamentals alongside reproducible software practices: packaging, data wrangling, visualisation, version control, and collaborative development with Git and GitHub.
2024/25 repository
Seminar on multilevel models · 2023/24 · TU Dortmund University
A graduate seminar covering linear and generalised linear multilevel models from both frequentist and Bayesian perspectives, with implementation in R using lme4 and brms.
Thesis supervision
2024/25 · TU Dortmund University
- Analysis of different initialization approaches for hyperparameter optimization with mini-batch stochastic gradient descent: a simulation study (Bachelor)
- Sensitivity analysis and performance evaluation of varying upper thresholds for discrete likelihoods using the Softmax-Gumbel Trick (Master)
2021/22 · Marburg University
- The influence of response scales on the knowledge gain of underlying cognitive mechanisms: the role of uncertainty and truth perception in the Truth Effect (Bachelor)
- Empirical test of core assumptions of the Referential Theory: influence of repetition on perceived coherence (Bachelor)
- Identification and testing of relevant psychological factors on truth judgments and the truth effect according to the Referential Theory (Bachelor)
2020/21 · Marburg University
- Truth Effect — the role of the response scale in truth effect designs with short delay (Bachelor)