Digital Behavioral Data
Winter semester 2026/27
This page provides an outline of the topics, contents, and assignments for the semester. Please note that the contents of the course sessions ( background information, slides, exercises and tutorials) will be updated continuously throughout the semester, with all changes being documented here.
| Session | Date and Time | Topic | Resources |
|---|---|---|---|
| 1 | 21.10.2026 13:15 |
What is DBD? & What is this course? |
|
| 2 | 28.10.2026 13:15 |
Investigating LLMs as Object and or Coder, Introducing our shared Dataset, Form Research Teams & Current CommSci AI Research |
|
| 3 | 04.11.2026 13:15 |
π¨ How our Data is collected: Ethics, Legality & Whoβs in the Sample. | |
| 4 | 11.11.2026 13:15 |
π¨ Describing Conversations: Features & Style Measures | |
| 5 | 18.11.2026 13:15 |
π¨ Building Typologies: Segmentation, Annotation & Reliability | |
| 6 | 25.11.2026 13:15 |
π¨ LLM-assisted Annotation & Validity | |
| 7 | 02.12.2026 13:15 |
π Preregistration & Colloquium: Pitching your Research Projects |
|
| 8 | 09.12.2026 13:15 |
π¨ Getting Started with your own Analysis | |
| 9 | 16.12.2026 13:15 |
π€ Guest Lecture: Prof. Johannes Breuer from the Center for Advanced Internet Studies on Chatbot Data Donations |
|
| - | - | π Christmas Break | |
| 10 | 13.01.2027 13:15 |
π Project Presentations: Intermediate Results | |
| 11 | 20.01.2027 13:15 |
π Project Presentations (cont.) / Group Work | |
| 12 | 27.01.2027 13:15 |
π¨ From Analysis to Write-Up: Reporting & Reproducibility | |
| 13 | 03.02.2027 13:15 |
Lessons learned, Q&A Your Term Papers |
The course follows a funnel: we start broad (what LLM chat data are and what they can tell us about communication), narrow into the methods needed for one shared corpus, commit to project designs in December, and spend the last third on your own analyses.
- π¨ Hands-on sessions working with R. Bring a laptop.
- π Design colloquium β every group pitches its project. Mandatory, but not graded.
- π Project presentations β graded, based on your intermediate results.
Sessions 3β8 each close with a short block in which you apply the sessionβs method to your own project, so that your project grows week by week rather than starting from scratch in January.
Copyright
This content is licensed under a GPL-3.0 License.