Both workshops started the same way: I noticed a gap in how the people around me were being trained, and I had the background to fill it.
Best Practices for Programming in R (2020–2023)
When I moved from software development into research, I found that practices the software industry treats as basic were largely missing from academic data work: naming conventions, version control, documentation, code review. Many researchers write analysis code without any formal programming training.
I didn't simply transfer industry standards. Software companies and research groups need different things: commercial software is a product that other teams will maintain, while research code has to produce results you can trust, rerun and share. So I kept the practices that serve those needs and left the rest out. For version control, for instance, I set out three options side by side (Git, dated file names, and a change log inside the script), each with its advantages and disadvantages. Learning Git just to keep track of versions is a lot to ask, and often more than research code needs. Making Git the only route would also have made version control look out of reach for less technically inclined scientists. So I offered two options alongside it that anyone could start using the same day.
Participants are invited to open a script of their own and compare it against each practice as we go, so every point lands on real code.
I also changed the workshop based on feedback. After my first delivery, comments came back from a couple of people who strongly resisted my advice to stop clearing the workspace at the top of a script. I kept the point and added a slide that answers the objection directly: you still need a clean workspace, you just get there a different way. The worked example that follows shows why it matters: run a colleague's script that starts by clearing the workspace, and your own work in progress is gone.
It covers RStudio projects and portable file paths, script structure and section headers, library declarations, version control, commenting, naming conventions, parameters instead of hard-coded values, and breaking code into functions.
I delivered it seven times, every one by invitation: R-Ladies Nijmegen, June 2020 · Open Science Community Nijmegen, October 2020 · R-Ladies Coventry, UK, December 2020 · R-Ladies Bergen, Norway, March 2021 · International Max Planck Research School, Nijmegen, as part of Key Practices for the Language Scientist, 2021, 2022 and 2023.
The slides are written to stand alone without a presenter and are free to use. → See the slides here
Shows: Training, facilitation & mentoring · Coding
Rhetoric for Cognitive Science Students (2015)
During my master's at Osnabrück University, a fellow Cognitive Science student and I noticed classmates whose presentations undersold the quality of their work. Together we designed and ran a workshop on presenting. It covered preparing material, building a narrative, using voice and speech, body language, handling stage fright, and giving and receiving feedback. The workshop ran two full days and included a variety of formats, group discussions, presentations, active participation, individual work, giving and receiving feedback from and to fellow students, as well as homework. The students started and ended with a presentation and made changes to the presentation as the class went along resulting in a final presentation.
Shows: Training, facilitation & mentoring · Communication across audiences