I'm an Invited Assistant Professor at ISPA. My work connects philosophy of science, research methods, and statistics: I teach across those areas, do quantitative consulting, and develop open-source software to help researchers build more reproducible and open research workflows. My academic background is in social psychology, with research on how adults perceive and treat children as a social category.
This GitHub profile is mostly the technical side of that work: R, reproducible research, Linux, research software, and small open-source tools. More recently, it also includes some experiments around LLM-assisted workflows.
I have a strong preference for software that is easy to inspect, easy to modify, and not much cleverer than it needs to be. If a small tool starts to look like a control room, I become suspicious.
A lot of my older work lives on GitLab, which is still the canonical home for several projects. I mirror more of it here now because GitHub makes the work easier to find and contribute to.
| Project | What it is |
|---|---|
| Pi Sych | A small package around Pi for longer LLM-assisted projects, with project state kept in ordinary files and bounded work handed to fresh contexts. Docs · GitLab |
| J's Pi Bakery | The home for my small Pi extensions and companion tools, including pi-pew-pew, pi-filler, pi-lease, pi-auch, and pi-tin. |
| tr.iatgen | An R package I co-developed with Michal Kouril for adapting iatgen-generated Qualtrics QSF files for non-English samples. Paper |
| NRB | An open collection of European Portuguese teaching materials on research methods, statistics, R, and scientific writing. |
| make-it-stop | A deliberately small static-site generator. Several of my own sites use it. The name is also a design requirement. GitLab |
- Research and data: R, statistical modelling, mixed and multilevel models, robust methods, visualisation, experimental and survey methods.
- Reproducible work: Quarto, R Markdown, Pandoc, LaTeX, Git, CI/CD, and analysis/reporting workflows that can be inspected as a whole.
- Software and systems: Linux, shell, TypeScript/Node.js, Make, research software, and small utilities built because I wanted something simpler—or wanted to understand how it worked.
- LLM-assisted work: experiments in making model-assisted projects easier to inspect, control, reproduce, and hand back to a human.
The code is only part of what I do. My academic work, publications, teaching, and supervision are in my CV, and I keep a broader selection of research, teaching, and software work on my Projects/Work page.
I also do independent consulting on research methods, statistics, and data analysis, especially when the hard part is deciding what analysis actually makes sense or turning a messy research problem into a reproducible workflow.
- Website — the short version of who I am and what I do
- CV — publications, teaching, academic work, and the formal record
- Projects — research, teaching, software, and assorted experiments
- Blog — notes on research, statistics, computers, AI, and things I probably could have kept to myself
- Services — consulting, tutoring, and workshops
- GitLab — canonical source for several projects
Most projects are small on purpose. A few are small because that is as far as I got.