AI for social science methods

Since the fall of 2025, I have been teaching a course about incorporating generative AI into social science research methods. The earlier versions of the course closely followed a special issue of Sociological Methods & Research that I co-guest-edited with Thomas Davidson (UNC Chapel Hill), which can be found here. Recent versions of the course have expanded beyond this special issue given the large amount of excellent research about using AI as a methodological tool.

These are my course’s recent programming-based activities:

  1. An activity to set up students with a local LLM model and provide exposure to R: HTML and R Markdown

  2. Activities about prompting:
  3. An activity about using LLMs to generate human-like data (and touching on “silicon sampling”):
  4. An activitiy about measurement and statistical bias: PDF and R Markdown

  5. An activity exploring model hallucinations and some exposure to NotebookLM: PDF and R Markdown