This course covers design and design-based inference for both surveys and randomized experiments. In covering the two domains, the course emphasizes the important and underappreciated relationship between design and analysis. The lectures go back and forth between discussing tools for surveys and experiments to highlight connections. Problem sets will emphasize both application and theory. A final project will require students to extend methods from the survey literature to the design and analysis of an experiment using real or simulated data.
Required skills. Students should have completed PS231A with at least a B or better. Students with less technical background will find the course pushes their capabilities.