Justin Li

Reflections on Teaching


I’m not sure when it became apparent to me that liked teaching. I remember helping other people with math and biology back in ninth grade, but I don’t think I had that much self-understanding at the time. I helped with some (in reality, a minimal amount) of peer tutoring in high school through the math center, where students can drop in and ask questions about their math work. Not too many students used that though, and I maybe helped three students in the two years I was there. Freshman year in college, I was a student in the Gateway Science Workshop (GSW) peer tutoring program at Northwestern. I think that was the first time I really woke up to the idea that I want to teach. I remember that my facilitator was really not very good at the whole facilitation thing, and often I would be explaining things to other people in the session, if other people were there at all. I stayed in the program because I was bored (the things boredom make you do!), and I stayed with the same facilitator because it was easy (the things comfort makes you do!). But by the last quarter of that year, I had made up enough of my mind to apply to be a GSW facilitator next year, which I got.

For the next three years, I was first a facilitator, then a senior facilitator, and finally just filling in sessions as necessary. This means I’ve been involved with the program for all four years of college, and of course I did two summers of CTY as well. In the next two weeks I will be a Graduate Student Instructor (GSI) at Michigan, my “fourth” year of teaching.

It is therefore unfortunate that this last year, if not more, of GSW was horrible.

By this statement I don’t mean that I grew to hate GSW, or indeed that GSW was something I had to tolerate. No, I still enjoyed every session of GSW; I just wasn’t convinced my students enjoyed it. When I think back over successive years of GSW, I can see that I have gotten complacent about teaching and have relaxed how I think about it. Let me give some examples.

Although the GSW facilitator orientation at the start of the year recommends doing ice breakers during the first session (or two) of each quarter, I gave up doing those around the middle of junior year. During the first year of GSW, I had a couple students follow me for the entire year. Since students can get priority to follow the same facilitator through the school year, this means that the students found what I said useful. I had some students join me the second quarter, who stayed for the final quarter too. But each quarter there were also new students, and so despite it being awkward for the people who I already know and who already know each other, I continued to do ice breakers. The second year, no one from my first quarter stayed, but a number from my second quarter stayed for the third. I didn’t intend to skip ice breakers – I remember coming prepared to do them, but there were only two students in the first session of the quarter, and it seemed silly to do games with just two people.

I could blame the decreasing number of students in GSW for that, but while that does contribute to the situation, it is my attitude which changed. Simply put, I became more focused on what an ideal session is like than on how to foster such an ideal session. For peer tutoring (and I think teaching in general), it’s best if the students talk to each other and exchange ideas. Not because there’s a barrier, however small, between facilitator and student, but also because students are much closer in mindset than the facilitator. Sharing how they approach problems allows everyone to reformulate the question, and perhaps find a perspective which helps them most easily understand. And therefore, in an ideal session (I remember I had one absolutely perfect session back in my first year), the facilitator mostly listens while the students teach each other, only stepping in if everyone is stuck.

My mistake is that I allowed my non-participation become status quo, without adapting to whether the students were discussion. This turns what should be a discussion into something worse than a lecture – it turns into a self study. I forgot that getting that initial engagement takes work, and an ice breaker – even if it seems silly – is part of that work. Instead of leading the students through problems until they have developed the confidence to discuss among themselves, I have just been letting them work.

I just finished Ken Bain’s What the Best College Teachers Do, and while it wasn’t a ten point plan to better teaching (which would have been bound to failure), I did get some insight into what I’ve been neglecting. The main thing I learned is that even though you may be going through the same motions, your relationship with you students depends as your attitude as well, and without that the same motion just doesn’t work. My personal example of that is offering to stay after the session to help. The first year, I’ve had students ask me more questions afterwards, and that had led me to know more about them and their background as well. Lately though, although I still make the same offer, no students take me up on it. This is because I haven’t gained their trust (note: this is different from them not trusting me enough). When I started GSW, I cared a lot more about what the students are thinking and how they are feeling about the material and the session itself. In my personal journal I have written short evaluations of the students after the first couple sessions, noting how they’re reacting to me, and how I can draw them out more. I actually cared about their answer to the customary “how was your week?” – it wasn’t just a greeting, but a semi-serious attempt to know more about the students. This mutual understanding and trust, although not reflected in the students’ attendance records or grades, paves the way to the autonomous learning I described.

The other big mistake I made is not leading them through the questions anymore. This applies whether or not that trust is established, although if the students are actively discussion the problem at hand it is more permissible to step back. What I realized is that talking them through the question (note: not through the solution) shows them more than how to arrive at the answer. It also shows them how to approach these types of questions, and what kind of thinking the student is supposed to have. After all, the material is only part of what we care about as teachers. Equally as important, if not more so, is that the students can think within the discipline. For computer science, for example, that means the ability to break down a complex procedure into simple steps, evaluating whether that solution is cost-effective, and so on. For Newtonian mechanics, that means seeing the forces acting on objects and knowing how the forces interact. Whether they can use big-O descriptions or integral calculus to find the answer is something else, but getting that thinking there takes precedence. If I lead the students through the question, letting them make mistakes and questioning them on why their solution does or does not work, I am changing ever so slightly the way they see the world.

The last mistake I made, this one somewhat less apparent but equally influential, is that I stopped thinking about what the students are learning. This seems silly, especially since GSW has all the worksheets made for us. But without a goal, without consciously trying to lead the students into the way of thinking, it is too easy to simply go for the answers. I am robbing the students of their opportunity to look at the larger questions. Again, it’s not the questions on the worksheet which are the most important, but the concepts behind it. The successful completion of the worksheet is merely a proxy to what the students learned. Bain brought up studies where the students can ace a Newtonian physics final, but still hold some naive, intuitive beliefs about motion. Part of it is certainly how the final was designed, but part of it (a larger part, I would argue) is that the teacher has not thought about how to bring it above just plugging and chugging.

If I had to summarize Bain’s book in one sentence, it’s this: The goal of teaching is for students to not just learn the material, but to understand it’s implications. This necessarily means changing (or at least adding to) how students think. Ways of thinking, unfortunately, are pretty resistant to change, and so to accomplish that the teacher needs to be prepared. I want to give three examples of good teaching, all of which I have personally experienced as a TA to the course.

The first is from CTY, where once a week during lunch my instructor will give a student a dollar if they can give the pronunciation and meaning of a word (limited to a collegiate dictionary) which he couldn’t spell. The result is that the day before and the day of, the students will spend a lot of their free time looking at a dictionary. Think about it: if a teacher’s job, in general, is to spark people’s desire to learn, then he has achieved this beyond measure. No children of twelve will willingly spend hours looking at a dictionary, but giving them this challenge drives them to do so. Whether they learned anything by looking at the dictionary is arguable, but it did inspire an educational fervor.

The second example is from compilers. The professor consistently brought up books and papers which were not included in the textbook. These books were not necessarily related to compilers, although they were all relevant to the topic under discussion at the time. This has the same effect as the spelling challenge – it draws the students deeper in to the subject, while making them aware of the larger context of what they’re studying. The best thing is that, at least for me, I really get the sense that the professor is really bringing up the books because he wants to invite the students to share his curiosity and interest in those subjects, and not just to impress students by how much he knows.

The last example is by the same professor. I mentioned that exams and most homework being just a proxy to measure how much a student understands. So the “final” for the compilers course (in addition to a working compiler, of course), instead of being a paper, was a “code walk”. That means the student/group comes in to meet the professor for an hour, during which the student walks the professor through their code. This is done in an informal fashion – it’s not a powerpoint presentation – but it allows the student to let the professor know, with the highest “bandwidth” possible, what they have learned. And although there are “points” awarded through the quarter for sections of a finished compiler, that merely influences the final grade. If a student did poorly before, but have improved as the quarter went on (each assignment is cumulative, and previous test cases are run again) with the final result of presenting a well organized code walk at the end, then they get a good grade. That is, the students are graded on what they understand, both about compilers and about programming in general, and not just whether they can answer questions about compilers in general.

So to put this into practice, I’m writing my own CTY robotics curriculum (like I did last year) with all this in mind. I will post the result here when I’m done. To close though, I want to pull a totally unrelated quote. I feel this should be how assignments given to students should be viewed, as Stanford University president John Hennessy said (as quoted in Thomas Friedman's Hot, Flat, and Crowded):

A series of great opportunities disguised as insoluble [read: intriguing] problems