In this two-day online course you will learn how to use Claude and Claude Code to write, debug, and understand code for empirical research in the social sciences—effectively, safely, and ethically.
What Is This Course About?
This two-day online short course introduces you to AI-assisted coding for empirical social science research using Claude and Claude Code. You will learn what large language models such as Claude are good at (and where they fall short), how to break a research problem into well-defined tasks a model can help you with, and how to use Claude's chat interface, Claude Code, and Claude skills in your own workflow. Throughout, the emphasis is on using these tools effectively, safely, and ethically—keeping a human in the loop, testing your code, and working in sandboxes so that AI accelerates your research without compromising its quality or integrity.
Learning Goals
After this course you will:
- Understand what tasks are well suited to AI tools such as Claude and which are not, and how to reason about task uncertainty.
- Be able to formulate a coding or data task clearly enough for a model to help you, and critically evaluate what it returns.
- Be comfortable using Claude's chat interface, Claude Code, and Claude skills for research coding.
- Know how to keep a human in the loop and set up review steps when coding with AI.
- Be able to build robust workflows using testing and sandboxes to catch errors and protect your data.
- Be aware of the ethical, safety, privacy, and reproducibility issues raised by using AI in empirical research, and know practical guidelines to address them.
Recommended Readings for the Course
- Korinek, A. (2023). Generative AI for Economic Research: Use Cases and Implications for Economists. Journal of Economic Literature, 61(4), 1281–1317. https://doi.org/10.1257/jel.20231736
- Dell'Acqua, F., McFowland, E. III, Mollick, E. R., Lifshitz-Assaf, H., Kellogg, K., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2023). Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality. Harvard Business School Working Paper 24-013. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4573321
- Anthropic (2025). Claude Code documentation and prompt-engineering guide. https://docs.claude.com
Assignments for the Course
Short course (2 days online, 8 hours total) | 1 ECTS point
Requirements to obtain credit:
- Attendance (participant has demonstrably attended 80% of all course activities)
- One formative in-class assignment (not graded): a short hands-on coding exercise in which you use Claude Code to write and test a small piece of research code, plus a short quiz on safe and ethical use.
Schedule
Day 1 – Foundations: what is a task, how do computers “think”, what Claude and other AI tools are good (and less good) at, and an introduction to Claude's chat functions, Claude Code, and Claude skills.
09.00–10.30h: Input session: tasks, how computers “think”, and what to outsource to AI (and what not to)
10.30–11.00h: Break
11.00–12.30h: Synchronous lab / tutorial: Claude chat, Claude Code, and Claude skills hands-on
12.30–13.30h: Lunch break
13.30–14.30h: Independent small-group exercise (guided)
14.30–15.00h: Office hour / Q&A
Day 2 – Building robust, responsible workflows: accounting for task uncertainty, setting up coding with humans in the loop, building robust systems with testing and sandboxes, and usage guidelines for safe and ethical use.
09.00–10.30h: Input session: task uncertainty, humans in the loop, and usage guidelines
10.30–11.00h: Break
11.00–12.30h: Synchronous lab / tutorial: testing and sandboxes for robust systems
12.30–13.30h: Lunch break
13.30–14.30h: Independent small-group exercise: apply the workflow to your own project
14.30–15.00h: Office hour / Q&A
In addition, students should plan for sufficient time to complete the assigned hands-on exercise and to install and set up the tools in advance.
Who Is Your Instructor?
Lena Janys is Professor of Econometrics at the University of Konstanz, working in theoretical and applied microeconometrics with a focus on health and labour economics. You can find out more about her work at sites.google.com/site/janyslena and wiwi.uni-konstanz.de/janys.
