Improving Classes with Live Student Feedback

One of the huge payoffs from having smart students that are engaged and thinking creatively is that they have great ideas for how I can make my courses better. Yesterday was the last meeting of the semester for my Economics of Human Capital in Latin America seminar, and I spent the last 15 minutes getting feedback on the course. I could have waited for their written course evaluations, but this way we can actually have a constructive dialog. Here’s what we came up with:

Substantive Content

  • They felt like they learned a lot about economics, policy, and empirical methods, but didn’t have a strong feel for the region itself. I should provide more background material (e.g., optional reading) on Latin America.
  • After the introductory week, weeks two and three are focused on schooling. My students choose their research topics right before the fourth week. Not surprisingly, topics are heavily weighted toward education. I need to do a better job during the first week previewing the later course content, but I’m also seriously considering making their research topics due a week later so they are exposed to a wider range of material.
  • Some students wanted to spend a week talking about how human capital influences (and is influenced by) poverty and inequality at a macro level. I used to have a class on this topic but dropped it. I might have to bring it back.
  • Other students wanted to know more about rural areas of Latin America and in particular the role of human capital in agriculture. I think most of the interesting work on agriculture in development economics is happening in Africa and Asia, but I should look more closely to see if there are good papers in Latin America.
  • For everything I add, something has to go, and I’m not sure what that would be. Maybe I should give them a menu of 15+ topics and have them vote for their favorite 13 at the beginning of the semester.

Empirical Methods

  • The literature we read mostly relies on data analysis methods they have never seen before. These include difference-in-differences, instrumental variables, regression discontinuity, and propensity score matching. I try to give short mini-lectures about these methods in class before they read articles that use them. Sometimes we run out of time and I email them some notes and review the method on the same day we discuss the corresponding paper. This is sub-optimal and I need to try harder to do the mini-lectures before they read the papers. That said, my students did appreciate the class’s balance of substance and method.
  • I should provide more references for the methods we talk about for those students that want to learn more.
  • In years past I’ve had some students suggest that I hand out a methods problem set at the beginning of the semester to ensure that they take some time outside class to mull the math behind the methods. This year a student suggested a short in-class quiz early in the semester for the same reasons. It would certainly help them later in the semester if they were more solid on the methods, so next year I’m going to try one of these ideas.


  • A staple of my seminars is the worksheet that every student fills out for every article they read during the semester. They always struggle at the beginning and then get the hang of it. To speed up this process, some students suggested that I walk through a few worksheets in class and hand out a few more examples of worksheets that are filled out well.
  • Almost every article we read has one key table or figure that summarizes the primary contribution of the paper. One of my students had the fantastic idea that I should ask them to identify this in the worksheets. Doing so would force them to spend quality time with the raw tables instead of trusting the authors’ textual interpretation of the results.
  • At the end of every worksheet, I ask them what they didn’t understand in the paper. Some of the students were frustrated that they didn’t always get answers to those questions in class. I tried to fix this last semester and the cure was worse than the disease. A better system is out there, but I still don’t know what it is.

Presentations and In-class Exercises

  • They all got a lot out of preparing and doing a presentation in class, and one student suggested that everyone should do this twice. I don’t want to spend more class-time on presentations, but if the individual presentations are shorter, this could work well.
  • They liked the presentations of things they hadn’t read such as background material or optional papers. We should do more of this.
  • In class I often have them suggest potential solutions to complicated social problems in a very small amount of time. If I pre-announce these exercises, they can think about them before class and have a better and deeper discussion during class.