I first taught research methods to master’s students at Columbia’s School of Public Health. The textbooks accessible to my students focused on descriptive research—measurement, surveys, that sort of thing. I thought, “This is important. Why didn’t I learn it in economics grad school?” But I also thought, “They’re missing the important material I did learn in grad school.” Those books limited causal research to a little bit on traditional experiments, leaving out the revolutions in natural experiments and causal inference with observational data.
So, I tried (and often failed) to make causal research accessible and relevant. When I moved to Baruch’s School of Public Affairs (City University of New York), my colleague Gregg Van Ryzin and I decided to write the book our students needed. After many drafts tested on our classes, we published Research Methods in Practice: Strategies for Description and Causation—now in its third edition.
I was especially proud that we went beyond “correlation is not causation” to show students like ours how to (responsibly!) learn causal effects from observational data. As I often told my students, “We won’t be randomly assigning families to divorce, but we still want to know how divorce affects kids.”
Most of my students aren’t researchers or data analysts. They’re managers, fundraisers, budgeters, political staffers… I wanted them to be savvy consumers of research. One student nailed it when they said, “This course is about: does the study say what it says it says?”
Over time, I realized my students could (and should!) be more than consumers of research. They need to commission and supervise data analysis, manage administrative data to enable useful analysis, and handle the constant flow of ‘what works and how well’ questions. Did our new staff schedule raise fall rates in the nursing home? Would new scheduling software cut no-shows? They don’t usually have time for a careful study—they need to think on their feet. I owe this shift in perspective to my students and to my practitioner co-instructor, Gregorio Morales.
Still, it was tough. I was teaching a mountain of concepts that demanded a lot of new ways of thinking. Students found the class relevant—but also overwhelming. For most, every week brought a brand-new analytical idea and lots of work. One said, “My brain hurts.”
I wondered, “Why do I have to cram so much into just 14 (or even 10) weeks?” You don’t need to be a “math person” to learn this material. But you do need time and practice—over years, not weeks. Just as Lynn Arthur Steen launched the Case for Quantitative Literacy, I’m launching the Case for Causal Literacy.