Thoughts on Teaching from the Department Chair

Encouraging Good Teaching

Research university like Berkeley have two major goals: creating knowledge and disseminating it. Both are important.

There are two big incentives to teach well. First, it's good for you emotionally. Second, it's good for your career.

Research can be frustrating and involves a lot of delayed gratification. Teaching brings instant gratification. You can usually tell when you've done well, and it's a good feeling. Working in a research lab is fine, but people in universities tend to last longer than people in research labs.

In this department, good teachers get promoted faster than bad teachers. Every three years, faculty members are evaluated for promotion. They are judged on three areas: research, teaching, and service. Research and teaching tend to be more important than service. Research is evaluated in terms of talks, papers, and the opinions of your colleagues. Teaching is evaluated in terms of ratings, comments from others, and self-evaluation (What did did you do to make your classes better?). Also, you can sometimes spot good teachers because the enrollment in their classes is unusually high.

Questions and Answers

Question: Is there a system to get feedback from GSIs about instructor?
Answer: No. GSIs can write letters, but there are no formal mechanisms.

Question: What ratio of research to teaching do you look for when hiring?
Answer: Seventy to eighty percent of the people we hire are assistant professors who may not have taught their own classes before. We look at the people they were TAs for, ask them about their teaching philosophy, and evaluate the quality of the talk they give when they are interviewed (although this is probably more polished than most lectures they will give in class). Basically, we look for a interest in teaching. As a tip when you are applying for a job, look at the list of courses at the school where you are being interviewed. Know which classes you could teach. Look at the research being done there. Know what they are doing and what you and they have in common. You will almost certainly be asked something along these lines, and it looks bad if you can't answer.

Question: What do you do if there are no classes in the catalog that fit you? Maybe that means they need you.
Answer: You should point out that they obviously need somebody in your field, but you need to find something that you could teach that they already have.

Question: Does the number of times you were a GSI help when you are trying to get hired?
Answer: It can help, but it depends on the university where you were a GSI and what they expected from their GSIs.

Question: What happens with the review process once you get tenure?
Answer: You still go through the same review process to get pay raises. At a university like Berkeley, almost everyone with tenure is a good researcher. The quality of your teaching may well cause you to stand out more than the quality of your research.

Question: Undergraduates do not always fill out the ratings seriously. How can you make students take them seriously?
Answer: I'm not sure. The best thing may be to throw out the obviously crazy answers.

Suggestion from a student: Solicit comments from student when a professor comes up for review. Alternatively, get student representatives to review professors.
Response: There are problems with ratings. They tend to be a little bit of a popularity contest, but the ratings in grad classes are filled out very seriously. There are more jerks in the lower division.

Personal Teaching Philosophy

The Spiral Method

People learn some things multiple times, with varying degrees of correctness. Giving them all of the details the first time won't help. Give them one layer first. It won't be right in the details, but it is a good basis for future teaching. This is also a good trick for research papers. Keep the introduction simple, and then add more layers in later sections.

Convey Intuition

Details can follow, but intuition should come first. This is one of those things that is better covered by a teacher than by a textbook.

Preparation, Preparation, Preparation

Thanks to Richard Karp for this. Make sure your presentation is consistent. Think about alternate ways of explaining everything, just in case students ask questions.

Corollary (Karp): If you make a mistake, admit it. You and the students should know you can make mistakes.

Side Remark: Overreliance on Technology

This is a personal philosophy. Don't become too reliant on technology. Edward Tufte has written a lot on how to convey an idea visually, using charts and tables and things like that, but he has also written an anti-PowerPoint tract (example).

Writing on the board can be good in some cases -- some information comes out at some times and other information comes out at other times.

PowerPoint and similar programs can make life too easy. When reusing slides, you might not feel the need to prepare or review. When you are on your own, you have to think on your feet. You rebuild understanding as a teacher.

Questions and Answers

Question: Students spend too much time taking notes. Isn't it better to get and annotate printouts of the lecture?
Answer: Annotating is good. Don't copy slavishly. The writing process should engage students.

Question: What about when a professor writes on a tablet that is projected onto the screen, and later everything is dumped to the web?
Answer: That works very well. It takes cool nerves, but it can be good.

Question: The campus is pushing the mentoring of GSIs. How did you do this, and what would you suggest?
Answer: Meet with GSIs. Drop in on sections. Watch them run discussions and give feedback.

Question: Do you have any hints for time management?
Answer: I wish I didn't, but I don't have time for it. Seriously, though, read one of those silly books. Of course, it's not always easy to follow their advice. For e-mail, see what you can deal with in two minutes. Classify the rest and deal with it later.

Question: How much time did you spend on teaching as compared to everything else, before you were the Chair?
Answer: Define teaching. Working with grad students? Lots. Pure classroom teaching? Maybe 25 percent. It varied with the class and how often I taught. The third time teaching was generally the charm. By the third time, you don't need as much preparation. After that, though, you can get bored. You need to change the class to keep it fresh for you.

Question: Do research advisors change their expectations when you are teaching?
Answer: They would be foolish not to. Still, you should be up front about it.

Mike's Observation: Be productive. Use TAing as a break from research.
Answer: It's best to TA in your second or third year, but that's not in line with Berkeley's reality.

Question: It's common wisdom that GSIs (especially 10-hour GSIs) work more than they are scheduled to. How do you bring reality into line with policy?
Answer: The party line: that's the professor's job. This is a hard question, and the Department can't do much. It can't micromanage every course.