“Night and Day”: Steve Hansen on Teaching

At a recent dinner, Steve Hansen, a friend of mine, said the difference between his current teaching and earlier teaching is “night and day,” partly due to this blog. I asked him to elaborate.

ROBERTS What is your teaching situation?

HANSEN I’ve been teaching at Peking University [in Beijing] in the Guanghua MBA program for the last two years. I teach courses in innovation (big company Clayton Christensen sort of stuff), entrepreneurship, and social responsibility/social enterprise. The classes usually consist of 30-50 students from all over the world.

My teaching at Peking U. is a part-time job. Full-time I run a social enterprise here in Beijing (Phonemica). Before this I was a professional market researcher working at a large company and I worked as a consultant for years. Although I had a little teaching experience early in my career (e.g. taught undergraduates when I was a grad student), I certainly did not have any special teacher training or background. I was hired at Peking U. on the basis of my professional experience. Thus when I started teaching I was to a large extent repeating to my students the experience I had had as an MBA student myself, years ago.

ROBERTS How did you teach this course at first?

HANSEN My teaching changed quite a bit in the first year. The first semester I did lecture and discussion. I didn’t like it. I felt it was a lot of work to prepare, and while it helped me personally a lot, by getting me to summarize my learnings from years of reading and work experience, I didn’t feel it gave the students a lot. Even though the course got mildly positive reviews, I felt the students weren’t engaged in a way that made me excited about going to teach.

The second semester I started making changes. First, I took to heart some advice I got to follow more of a case study approach. The experience I gave the students in the beginning of that semester was fairly standard “case study” method made famous by Harvard. Here‘s a mini-documentary of how Harvard expects case study method to work. The professor asks pointed questions from the case and expects the students to have answers and defend them.

ROBERTS What caused you to change how you teach it?

HANSEN I like the case study method in theory. I like it a lot. I think by mimicking real life situations and summarizing the many facts that synthesize a business manager’s knowledge and must be put to use in making business decisions, the case study method has the potential to help students learn in a way that is useful: giving them the skills to make real-world decisions, and doing it faster than they could get the same experience by going out and working. As far as activities that can be done in an academic environment go, the case study is pretty darned close to “learning by doing”.

However, I found my classroom reality was different from the ideal portrayed in that Harvard Youtube video, in which a professor says, “I have been amazed… at the consistent level of high preparation”.

In my reality, students often didn’t come prepared. Sometimes they would say, “Sorry, I didn’t have time to read the case” — despite significant portions of the grade being dependent on in-class participation. And even if they were prepared, to some extent, often it wouldn’t be to the extent that I had hoped, where they were able to argue the intricacies of a position and defend it. Instead they might give a cursory answer, with a weak defense, and much of the class would be spent with me pushing them to elaborate their answers through Socratic questioning, lecturing, and so on.

So I was still doing a lot of the talking, and really leading the class. It was better than my first semester, but it wasn’t meeting my standards of what a real learning environment should be: student-led because they’re motivated to learn.

It was about at that time that I must have read some of the articles and ideas you published about “learn by doing”. Those rang true to me. Learning by doing is something I’ve always believed in and tried to practice, and I realized that my students still weren’t doing it.

So later during that semester I started experimenting. My first experiment, I’m sorry to say, was a complete failure. My approach for a few classes was to have students simply teach the class. I would assign the next class’s case study to a group. They would read the case and come up with main points. I would meet with them and review what they were going to do. Then they went in and tried to teach their fellow students.

In retrospect it seems obvious that this was wrong. As business school students the “doing” that I wanted them to do shouldn’t have been to “do teaching” but to “do business” as simulated in the case. When the students were teaching the case they were (again naturally, in retrospect) worse than I was. They would lecture rather than ask. They would miss key issues. And of course they had the same problem I’d noticed myself: other students weren’t prepared enough, even though the lecturing group was extremely prepared.

After enough experimentation to realize this wasn’t working, I cut it off and went back to my semi-successful case study method.

But the failure had given me an idea. I saw that students who were forced to present were MUCH more engaged than those who sat and listened. Why not, I thought, just make everyone present? It would kill several birds with one stone. First, I’d noticed that students’ presentation skills needed a lot of work. They had trouble being concise and answering the critical questions. Having to present would give them practice on this critical skill. Second, they’d be engaged and motivated — at least for the part where they were presenting. Third, by divide-and-conquer of the case itself, we’d get far deeper into the details and (sometimes) quantitative analysis of the cases.

And so that’s what I started doing this year. It has worked even better than I imagined. Not only are students engaged for their part of the lecture and discussion; the preparation seems to spill over into the whole case, so most students really know what’s going on. The presentations get better and better as the semester progresses. In contrast to the past, now I’m having to shut down vigorous discussions in the interest of time, rather than ask pointed questions to stimulate interest. And best of all, this environment satisfies my own desire of creating a place where students want to learn. I work less, and the students learn more. That’s the part that seems “night and day” to me.

For anyone who’s curious, there are some details I’ve worked on to make the approach better, at least for my particular situation. Here’s the nitty gritty:

I divide the class into random groups of 4-5 students. That group is together for the semester, and for each class they’re tasked with presenting on one part of the case. I assign that presentation the week before, and it’s often a question that forces them to take a position. Often, too, I assign the same question to two different groups. This has the effect of creating a sort of debate between the groups, because they almost never come up with the same answer to a question.

I require VERY short presentations. Usually 2 minutes. Sometimes 3 or 4 if the topic is very complex. I allow any member of the group to present. Although I encourage them all to take the opportunity to practice presenting, I don’t force the issue.

All the students in the group get the same grade, but with the following “peer evaluation” element that I adapted from someone else (i read about it online but have now lost track of where). Peer evaluation means that your fellow students give you a grade at the end of the semester. If your fellow students all give you 100%, then you get 100% of the group grade. That’s the usual case. On the other hand, if all your fellow students give you 50%, that probably means you didn’t do much and you get only 50% of the group grade — a guaranteed failure for that portion of the class. In even more detail, what I do is take the average of the peer scores, but not including the lowest score. This virtually eliminates the possibility that one enemy classmate can sabotage a groupmate’s grade.

Overall I’ve found that the peer evaluation mostly eliminates the freeloader problem that general plagues groupwork. In cases where someone tries to freeload, they simply get a bad grade, so the other groupmates aren’t really bothered about it.

At least as importantly, this entire system, once explained, makes my grading job much easier.

I also like the “group is responsible for the presentation” approach because it allows some flexibility. There’s always going to be a week during a semester that a student simply gets swamped with other things and doesn’t have time to prepare. Under the group presentation regime, the busy students can beg groupmates to cover for them, and then return the favor another week. Would it be better if everyone were always prepared for every class? Well, maybe, but that doesn’t really seem realistic.

ROBERTS What did you read that influenced you?

HANSEN I read all your posts about education and teaching. It’s hard to say any one post made me change the way I teach. It’s more that your posts about teaching, like your posts about diet etc., often help me consider what my goals are and how I can do low-risk experiments to find better ways to reach those goals.

As for specifics…

I remember reading this, for example, that helped me remember to think about my students’ goals rather than my own for them (not that I think I should have no influence on their goals — more that I should keep in mind they are autonomous agents about whom I know little and should learn more).

I still grade, but I’ve experimented with “grading less” I guess you could call it, partly inspired by discussion like this.

I was probably most influenced/inspired by the Halmos paper (also here) you mentioned. I was a math major as an undergraduate, so I liked the subject matter. I was always interested in the kind of simple statement of problems and choices that he describes in that paper, not just in math, but in business and elsewhere. I really enjoy simple problem statements that force one to go through complicated analyses and yet still reduce everything back to a simple statement of resolution. Business at its most interesting can be quite like that. You have to make decisions: fund this project or don’t. Build this product or build that one. What do you do?

Even more than enjoying the problems, though, I completely agree that the hard part is making students ask questions. It’s not actually solving the problems that’s hard; it’s figuring out what the problems are in the first place. After reading his paper, I started doing some searching and found others who were experimenting with teaching techniques. It would take me a while to re-find those people, but I definitely gleaned ideas from others. These helped shape my own experimentation.