Teaching Applied Econometrics B. Wade Brorsen Anyone can be a good teacher and there are many different ways to teach that will work. Even so, there are a number of general principles that need to be followed. The objective of this paper is to list a few of the things that I wish I would have known before I began teaching. I have organized them into the following categories: (i) general principles of teaching, (ii) items to remember when teaching applied econometrics to graduate students, and (iii) how to get term papers that are not junk. General Principles of Teaching When I began my career at Purdue University, I was required to take an eightweek seminar series on teaching. In eight weeks, the course only made one point, so let s begin with it as the main point since it must be important. To teach, you must first decide what it is that you want to teach. This means that you must begin with student objectives. Once objectives are stated, everything else follows. In my econometrics class, I want students to be able to do research using econometrics. I then have more detailed objectives and my course is structured to see that students accomplish these objectives. So teaching involves telling students what you want them to know or to do, showing them how to do it, letting them practice doing it, and then testing to see if they can do it. These straightforward principles are what teaching is about. When teaching a new class, guessing the level of the students is one of the hardest tasks. If you guess too low, students are offended that you think they are stupid and if you guess too high, students will be lost. If one of the lost students wanders into the Dean s office to complain, then you will quickly find out that teaching really does matter. - 1 -
Most new professors overestimate student s abilities. There is an old story about the new professor that did not know what to teach the second day of class because he had covered everything on the first day. Davidson and Ambrose spend chapters arguing not to overdo it. One of my colleagues recommends giving a quiz the first day to determine what students already know. It is easy to forget how little you knew at one time and how much time it took to grasp concepts that now seem so simple. I must constantly guard against trying to teach too much. This less is more philosophy leads to two corollaries: (i) if it is not important enough to say out loud, then do not put it on the overhead, and (ii) if it is not important enough to be on the test then do not cover it in class. One question that often comes up is what is the best method to use in giving a lecture? For a large lecture class, there are certainly advantages to Power Point because it will let you produce visuals that are clear and large enough that everyone in the class can read them. Students seem to be ambivalent about Power Point. I use a marker board and hardly ever use Power Point in class. Writing on the board lets students see derivations develop and keeps me from going too fast, which I have a tendency to do. I do hand out my notes, but I act as if students do not have them. In fact many students choose to not bring my notes to class and to write their own notes. I hand out my notes because some students want them. I see no reason to hide information. With the many equations that are used in econometrics, giving out my notes reduces opportunities for students to make a mistake copying from the board. You will also make mistakes. Making mistakes used to bother me a lot. I finally have accepted the fact that I am not perfect and make mistakes. Dale Carnegie said that if you make a mistake that you should admit it loudly and profusely and I can give no better - 2 -
advice than he did. So expect to make mistakes and be gracious to the students or others who point them out. Learning research shows that providing quick feedback on assignments increases learning. Make it a priority to get all assignments (including tests, but not term papers) graded and turned back to students the next class period. For one thing, it shows that you care. If you cannot give students the feeling that you care about them, then you could easily be replaced by a video. Professors are evaluated in part based upon student course evaluations. The conventional wisdom is that the best way to get high ratings is to ask little effort from students and to give high grades. The research does show a positive correlation between course ratings and grades as well as a negative correlation between effort and ratings, but the correlation is not very high. The reason the correlation is so low is that students care about many things other than grades and many of them even appreciate learning. If you want to get high course evaluations, I will give some advice, although some of this is my opinion and is not backed up by research. The number one factor is fairness. Because ratings are positively skewed and graduate classes are small, even one student who gives the lowest possible ratings in every category can wreck your ratings. This is another reason to make sure that you have clear student objectives that communicate exactly what is expected. But, you need to stick with your objectives. High expectations are fine. Students will typically meet high expectations as long as they are communicated. But, surprises are not a good thing. Our macroeconomics instructor recently gave a midterm exam with a question on it that he had not covered the material needed to answer it. How do you think students evaluated him? - 3 -
In undergraduate courses, being enthusiastic, entertaining, and funny are ways to be popular. This is not as critical in a graduate class since most students want to be there, but it is still a good idea. While I still find it odd, there are students who think econometrics is boring. Davidson and Ambrose say to show up five minutes early and visit with your students. This is clearly practical in a graduate class and I have found it helpful. It is also consistent with showing that you care. Finally, I would say to read about five books on teaching, because the best way to get good ratings is to actually be a good teacher. Teaching Applied Econometrics to Graduate Students While most of the principles of teaching graduate students are the same as teaching undergraduates there are differences. One of the most important differences is that graduate classes have students for whom English is not a first language. While all students will be able to read English, there will be some that get by the TOEFL with very poor listening skills. My rule is that any announcement made in class that is only said out loud and not written on the board does not count. Questions as well as responses to questions often need to be written on the board as well as spoken. Further, reading assignments need to be available on all topics. The assignments can be articles, books, or handouts. If you want students to learn applied econometrics then students will need to practice doing applied econometrics. The outline for covering a topic goes something like this: (i) give an overall picture of the method (ii) explain where the method is used so that students will be motivated to pay attention, (iii) provide the mathematics for the method, (iv) give the simplest example possible that illustrates the method, (v) give a homework - 4 -
that illustrates the method s use on a real research problem, and (vi) give a test question that determines whether the student can apply the method in a new situation. The key is making methods seem simple rather than making them seem fancy. Teaching is not the place to show off about how fast you can derive some involved mathematical expression. In my class, students learn more by doing the homework than they do by listening to my lecture. Learning to use statistical software is of course part of an applied econometrics course. Yes, I do spend class time on computer programming. Even though a class is applied econometrics, students still need to be able to do some mathematical derivations. They need complete understanding of properties such as unbiasedness and consistency. An understanding of these concepts is needed to select among methods as well as to have any hope of reading an article in Econometrica. Teaching Writing Since the objective of my course is for students to be able to use econometrics in research, my course requires a term paper. I have heard a colleague say that he used to assign a term paper, but all he got was junk so he quit. If you do not give students some help, then you can expect to get nothing but junk. I have attached a paper as an appendix, which gives some instructions on writing a journal-style paper. Reading the paper gives the students some help, but they will need more than reading a paper. Writing assignments need to be done in sections. The topic must be selected early in the semester. I allow students to use their dissertation research, since I am interested in getting them done and out the door. It is also imperative that students be required to write a rough draft and to revise their paper in response to detailed comments. There is no - 5 -
better way to teach how to do research using econometrics than to have students do real research using econometrics. References Davidson, C.I., and S.A. Ambrose. 1994. The New Professor s Handbook: A Guide to Teaching and Research in Engineering and Science. Boston, MA: Anker Publishing Co. (Author s note: This is not a great book as a whole, but it does have a few good ideas). - 6 -