This semester I have about 140 students in my econometrics class. We meet twice a week in a big lecture hall, and while I try to make it as interactive as possible, it’s not nearly as personal as the small (10-20 student) classes I teach. It’s also a lousy environment for students to learn how to actually analyze data. This is something they have to learn by doing, and most folks who teach data analysis expect students to acquire the skill on their own by working through projects or problem sets. I think in-person support can greatly accelerate this kind of learning, especially in an introductory class.

Last year I planned non-traditional lab-style discussion sections that addressed both of these issues. My teaching assistants (TA’s) would get to know their students so they could provide exactly the support each student needs. The TA’s would get feedback on the class from students that might not get shared with me during lecture or even in an anonymous survey (e.g., grumbling about the date of the midterm). And when students have a personal relationship with a teacher, they work harder and are more likely to attend class.

During these discussion sections, students work in pairs and use the statistical methods they are learning to answer substantive questions with real data on their own computers. The TA’s circulate around the room helping when students get stuck, checking in with everyone, and providing helpful advice. They never tell students what to type to get the answer and move on to the next problem. They rarely address the class as a whole group.

Last year, it took several weeks for this vision to become reality, and by that point half my students had given up on attending sections. The remaining students went on to have a positive experience, so this year I wanted to do everything I could to get it right from Day One.

First, I shortened the assignment to one that could be completed during the 50 minute section. The steps that got cut went into a part of the assignment called “For Further Study”. It’s amazing how frustrated my students (and my TA’s) got last year when they couldn’t finish during class.

Each of my four TA’s runs two discussion sections. This week, I attended each TA’s first section. I made sure they briefly (~2 minutes) introduced themselves and the structure of the lab, got students organized into pairs, and then got out of the way. This is way harder than it sounds since for most of my TA’s, their only experience in a discussion section is standing at the front and either lecturing or doing problems on the board.

Third, my teaching assistants each attended an extra section. This meant we had two teachers in every section–Either me and a TA or two TA’s. This was super-helpful the first week when students are using the data analysis software (Stata) for the first time. In this context, it’s very easy to get stuck and have no idea how to get unstuck. As the students learn problem-solving strategies (e.g., how to read the online help), they need less support.

So was it successful? I think so. I really enjoyed working with individual students, and I know it went better than last year. On Wednesday I start my weekly lunches with subsets of the class, and it will be interesting to see what they thought. I’ll also be tracking discussion section attendance–Since sections are optional, this might be the most objective measure of how they’re going.