June is the time of year when most faculty are just settling into a summer away from teaching. Not me though–after submitting my spring grades I had about 2 weeks of catching up on everything I put off during the semester, and now I’m back at it teaching econometrics online in Yale Summer Session.
The class is an intensive five weeks as my students learn a whole semester’s worth of material by watching many hours of recorded video, reading the textbook, and doing problem sets. They also see me in a live video chat room (imagine a big Google Hangout) three times per week. I have eight students and they get as much one on one time as they want–It’s much more like a small seminar than it is like a massive open online class.
Last summer I incorporated a whole load of new technology (e.g., Zoom and Canvas and I reported often on how it went. This year, the technology is much more stable, but I’m incorporating some of the lessons I learned teaching this class as a big lecture in the fall.
The class’s full name is “Econometrics and Data Analysis I” and historically it’s been more about the former than the latter. That is, students learn the mathematical foundations behind the methods and how to think about the methods, but they don’t get enough practice actually using the methods to analyze data. This summer my students will be loading up data sets, creating variables and describing distributions for a problem set that’s due on Friday of the first week. Last year they didn’t get their hands dirty until after the midterm, and that was too late.
The bigger change is a complete reboot of what I called “the big empirical project.” I used to give everyone the same large data set and a carefully structured set of steps that they would follow to answer several substantive questions. If they happened to not care about the topic (most recently the effectiveness of drug-eluting stents), it was kind of a slog. This term each student has to come up with their own substantive questions and answer them with appropriate data. I’ve put together a few candidate data sets, but already one student has run off to analyze Lahman’s incredibly comprehensive baseball database.
The keys to a successful creative project are getting everyone started early with intermediate deadlines, and providing lots of support along the way, since at the beginning they won’t know much about the tools they will use later on. It’s going to be a learning experience for everyone involved, but I couldn’t be more excited to see what these students come up with.