The Art of TAing: A Guide for Tutorial Teaching Assistants

Preface

This article prompted a writer from Nature to interview me and include my experience in the article Education: Time to teach. I think this quote from the interview cemented my inclusion: "I was a big critic of teaching assistants when I was an undergraduate. I thought I'd better not be one of the bad ones."

Introduction

This is an overview of my insights into how to succeed as a tutorial teaching assistant (TA) in a university. I was a tutorial TA eight times for three different courses in electrical and computer engineering at the University of Toronto. I won three departmental TA awards, so I had some success. Although I TAed electrical engineering courses, I expect that this material is applicable to any problem-solving oriented tutorial in engineering or the sciences. Below, I use some examples from electronics. If you don't understand them, simply ignore them as they are not critical.

Why Teach?

I found two good reasons to teach tutorials. First, I enjoyed it. I liked imparting my knowledge to students and helping them learn. I am amazed at how quickly students absorb knowledge and I am often impressed with the depth and intelligence of their questions. The second reason to teach tutorials is to learn the material. Of course, you probably know the material, but if you really want to know it, then you should teach it. In fact, teaching is a good test of understanding. If you can teach it, you understand it. Sadly, it's not necessarily true that if you understand a topic, you can teach it. I'm sure you have seen this proved by seemingly intelligent professors who couldn't teach a kangaroo how to jump. It takes sincere effort to construct a concept in a way that's understandable to others.

If you're still not convinced, Nobel-prize-winning physicist Richard Feynman has another reason:

In any thinking process there are moments when everything is going good and you've got wonderful ideas. Teaching is an interruption, and so it's the greatest pain in the neck in the world. And then there are the longer periods of time when not much is coming to you. You're not getting any ideas, and if you're doing nothing at all, it drives you nuts! You can't even say “I'm teaching my class”.

...

The questions of students are often the source of new research. They often ask profound questions that I've thought about at times and then given up on, so to speak, for a while. It wouldn't do me any harm to think about them again and see if I can go any further now. The students may not be able to see the thing I want to answer, or the subtleties I want to think about, but they remind me of a problem by asking questions in the neighborhood of that problem. It's not so easy to remind yourself of these things.

So I find that teaching and the students keep life going, and I would never accept any position in which somebody has invented a happy situation for me where I don't have to teach. Never.

—Richard Feynman, in Surely You're Joking, Mr Feynman!

(As a side note, it always helps to quote a Nobel-prize winning physicist when trying to make a point.)

What is a Tutorial?

In engineering courses at the University of Toronto, the lectures are taught by the professor (instructor) and describe the concepts and main material of the course. The tutorials are run by the TAs, who are usually graduate students, and demonstrate problem solving using the techniques taught in the lectures and answer students' questions regarding homework and past quizzes. In the courses I TAed, the tutorials were weekly 50 minute sessions or biweekly one hour and 50 minute sessions. The tutorials were held in rooms with chalkboards and the number of students ranged from 20 to 60.

The point of a tutorial is to help students learn the course material through solving problems. Helping you learn the techniques for helping students learn is the focus of this article.

Like Your Students

The first rule of teaching is that you must like your students. In fact, how can you not? You certainly were a student at some point in time. Did you not like yourself at the time? If you manage to hate yourself, hate teaching or hate students, please do yourself and your potential students a favour and ask for an appointment that minimizes your interaction with students.

Some TAs start off with a positive attitude towards students but become disenchanted over time because they closely interact with the enthusiasm-sucking students and only loosely interact with the (large majority of) enjoyable students. As a TA, much of your interaction with students will be regarding complaints about quiz and midterm marking and perhaps dealing with instances of cheating. Remember that these students are not a good sampling of the student population, and that most students are honest and work hard to learn the material. Also, you can increase your interaction with good students by simply engaging random students in conversation before the tutorial. A random sampling of students will yield mostly good students. And if you happen to chat with a normally bad student, maybe you'll win him over with your familiarity and pleasant attitude.

Respect Your Students

Do not yell at, curse at, insult, belittle, tease, or otherwise embarrass your students. The main reason you shouldn't act like this is that it's just plain wrong. Another reason you shouldn't do this is that even if you feel that a student deserves a dressing down, any emotion the student feels will be felt by the entire class. Students identify very closely with each other for the obvious reason that they are classmates. Attacking an individual student is like attacking every student. Of course, this doesn't mean that you have to be a pushover. Feel free to (politely) stand your ground if your students are demanding something that is unreasonable, but never attack!

Dealing with problem students is something I rarely had to do, but the phrase please see me after class is very powerful and you can usually pressure a student to modify his behaviour. Then you can deal with problem later by chatting one-on-one with the student. Also, remember that you can defer to the course instructor and simply ask students to take up an issue with her.

Respecting your students goes further than simply not being mean to them. If you respect your students you will also do your best to teach them. If you are reading this article, you probably care about doing your best, so you're on the right road.

Learn From The Best

I was somewhat lucky in this regard. In my first tutorial TA assignment during my masters degree, the course was taught by Prof. Ali Sheikholeslami, who later became my Ph.D. advisor. My fellow tutorial TA was Shahriar Mirabbasi, then a Ph.D. student and now a professor. Both have won multiple teaching awards. Ali is, without a doubt, the best teacher of electronics I have ever seen. He has an understanding and fluidity with the material that I didn't know was possible. Shahriar was the god of TAing. He basically swept the TA awards during his tenure. In fact, I heard that the department instituted multiple TA awards because Shahriar kept winning every year. It's hard to estimate their influence on me, but they both had a huge effect on my learning and enjoyment of being a TA.

So position yourself to succeed. Ask a professor you admire or hear is a good teacher if you can be a tutorial TA for her. Most professors prefer TAs who are enthusiastic about teaching. When I wanted to get some experience TAing outside electronics, I asked to TA Prof. Glenn Gulak's computer architecture course, because he was yet another teaching-award-winning professor from whom I knew I would learn a lot. And I did.

Be Patient

Be patient with yourself. This guide is quite comprehensive, but please don't think that you have to incorporate all these elements in your first ever tutorial. You will naturally have some difficulties in the beginning as you gain your bearings. I was a decent public speaker before my first ever tutorial, but I found it almost impossible to solve a simple linear equation in front of the class. My normal abilities eluded me while standing there! Luckily, I had brought notes in case my memory failed. As you become comfortable and encounter different situations, you can try a few of the techniques in this guide. Also, don't worry about making mistakes. I certainly made tons of mistakes and I am certain you will too. Learn from them and improve.

Be Prepared

If you remember only one suggestion from this article, it is that you should be prepared for any tutorial you teach. Over the years I averaged about two hours of preparation for every hour of teaching. This was true even for courses that I taught multiple times. And even on the occasions where I had to repeat a one-hour tutorial a few days apart, I usually spent another 20 minutes re-preparing before the second tutorial. The few times that I didn't properly prepare, I regretted it. Standing up in front of a class and teaching requires you to have fluency with the material.

My personal rule-of-thumb was that I aimed to understand the material one level deeper than what I was teaching. The more advanced the material and the newer you are to the area you are teaching, the more difficult it is to go one level deeper, but go as deep as you can in the time you have allotted for preparation. Remember, the students in engineering and the sciences are selected for their cleverness and perseverance and will ask intelligent and probing questions. It is your job to do your best in preparing to answer those questions.

You should be prepared in two senses. In the first sense you should be aware of the material that is being taught in the course and be able to explain all key concepts. Second, you should be prepared in the sense that you have worked out the solutions to all examples you will be presenting to the class. Skimming the answer book doesn't count!

A good way to learn new ways of looking at and understanding problems and course material is by asking your fellow TAs and graduate students. My friend Marcus was a great sounding board for ideas when I was stumped. He often had a different way of looking at electronics, perhaps based on the fact that, in addition to electronics, he's knowledgeable about how everything (yes I mean everything) mechanical works, whereas I prefer that the only moving things are electrons.

Write Neatly on the Chalkboard

Don't use overhead slides and a projector. I prefer to use chalk so that the students can see the progression of solutions. Write neatly on the chalkboard and take a moment to plan how you will use the board space. Will you keep a diagram throughout a problem? What portions of the board can you erase?

At the end of your first few tutorials, go to the back of the room and look at what's left on the board. Can you read it? Would you be embarrassed if someone photographed it? Change your approach to fix any problems you see.

Erase the Chalkboard When You're Done

This is simply a nice thing to do for the next person using the board. Do it even if the person before you didn't erase the board for you. And give up your seat to senior citizens and pregnant women on public transit.

Admit Ignorance

Even with the best preparation, you will be occasionally stumped by a student's question or unable to reason about something that you know is correct. That's okay; feel free to admit it. If you have done a good job on other parts of the tutorial, students will accept your admission of ignorance. This is a good opportunity to research the specific topic and bring in a solution to the next tutorial.

Defer Inappropriate or Inopportune Questions

There are no stupid questions, right? In the sense that you should never belittle students or their questions, this is true. Always accept students' questions at face value and never tease or criticize their questions. There are questions that are too basic or off-topic to take class time to answer. It's okay to say there's not enough time to answer that question effectively and then offer to answer it outside of class time.

Quiet the Class

Sometimes a critical mass of students gets rambunctious and talkative and you have to find a way to silence them so you can continue the tutorial effectively. I tried several methods of silencing including talking louder and louder, clapping, and banging the table. None of these methods worked repeatedly. I found only one method is infallible: silence. If I am speaking and the class's background murmur approaches a roar, or a single student is being disruptive, I simply stop talking and wait. I do not start speaking until the class has reached the exact volume that I am willing to tolerate. When you try this, it may take five seconds of waiting or it may take one minute. However long it takes, stick it out! This method has never failed me. Keep in mind, that one minute of silence can seem more like an hour while standing in front of a class of 50 students. I avoid feeling uncomfortable by taking the opportunity to watch the students. You can identify which students notice your silence immediately. You can also identify students who are completely oblivious to the silence around them and who keep talking until one of their friends shuts them up. Learn about all the fun and weird behaviours of humans in groups. It's like a zoo, but without the fur and cages (please, no photographs).

Use this power wisely. Absolute silence is not the goal. The goal is a level of background noise below the distraction threshold for students who are interested in listening. A few students chatting quietly is okay. If students are potentially helping each other to learn without interfering with the tutorial, why stop them?

Ask Questions of the Class

One way to occasionally punctuate the material you are teaching is to ask questions of the class. Note that I used the word occasionally. There is no more annoying method of teaching than continuously asking the class hundreds of pointless questions. The art of asking the class good questions has two parts. First, it is important to frame a good question. Good candidate material for a question is something that initially seems like a paradox. For example, you may use Miller's theorem in calculating the high-frequency pole in a two-pole amplifier. When this gives the wrong answer, ask why. Was Miller wrong? Of course, the class will find that you used the dc gain of the amplifier when applying Miller's theorem, but that gain isn't valid at high frequencies. So, the class learns by example that Miller's theorem is only useful for calculating the lowest-frequency pole. Sometimes you can plan these situations, sometimes they will just appear. Take advantage of them.

The second part of asking a good question is leaving enough time for the students to think. If you only give five seconds, they won't think about it. Just like silencing the class, you may begin to feel uncomfortable for pauses longer than 10 seconds. Persevere! Sometimes, one student or a small number may be able to answer very quickly, while the remaining students are still baffled. It's okay to ask the faster students to wait while the rest of the class thinks. If you feel students aren't thinking about the problem while you wait, you can let them know that you will randomly choose students to answer the question.

This brings us to another use of questioning the class, which is to get a feel for how well the class understands the problems you are solving. One way to ask this type of question is to pick a random student and ask a question to see whether he understands the current problem. Feel free to tell the class ahead of time that you will pick random students for questions, so that they are not caught off guard. Do not use this as an opportunity to ridicule students who don't know the answer. A student who says “I don't know” or gives an incorrect answer is giving you valuable information. This student's lack of understanding is probably true of other students too. So treat the student with respect, as you would any person who is a source of valuable information. Thank him for his honesty and address the misunderstanding.

Some students are so shy that they may freeze in the face of a direct question. Give the student a bit of time to calm his nerves and answer. This can become a confidence-building exercise for the student. On the other hand, if no answer seems forthcoming, it's okay for you to let him off the hook and to move on to another student.

Don't Hold Grudges

Sometimes students will create problems by arriving late, disrupting the class by talking, or some other creative method of distraction. Deal with the problem on an individual basis but invite the student to join your tutorial in the future without further censure. It's not fair for a student to pay the price over and over again for a single misdeed. Do not hold grudges.

Motivate the Material

Spend a small amount of time motivating the material you will teach in a session. For example, when introducing the concept of power consumption of digital circuits, I found it worthwhile connecting the material to current problem of power consumption in microprocessors. Since all students use and own computers, they are familiar with the problem high power consumption and heat dissipation of today's microprocessors. Furthermore, they have heard about this problem in the non-technical media, so it's almost like what they are studying is like, cool, in a way.

The Questions Are More Important Than the Answers

.. or problems are more important than the solutions. This point is critical for two reasons. First, it sounds zen-like. Second, and less important, it's true.

An interesting thing will happen when you step in front of the class and explain a problem. You will quickly run through the problem description and head straight to the solution. This is especially true when the problem is an assigned homework problem from the textbook. After all, the students have already seen the problem. Well, they haven't! I find that few students keep up with homework. Most do homework problems in batches, usually with the first big batch right before the midterm, and the second big batch right before the final. Regardless of whether the students are keeping up or not, a beautifully explained and well-thought out explanation of the answer is worth absolutely nothing if the students don't understand the question. So take the time to address the question. Make sure that the students understand what is being asked. Then, and only then, proceed to the solution. In many cases, exploring the question takes most of the cognitive energy. Once the question is understood, the solution is merely a quick set of computations.

Tell a Story

For a moment, pretend that you are human and that the students are also human. Tell a story. Tell the students about how you flunked a quiz in the course when you were an undergraduate—yes, the same course you are now TAing. Tell the students some interesting background story about some of the engineers in the field: e.g. Intel was started partly because William Shockley, a co-inventor of the transistor, was a jerk; Claude Shannon was a joker; or, you can always resort to telling a Feynman story. Bonus points if you can relate the story to the material.

Talk to the Instructor Regularly

It's important to know what material is being covered in the lectures and at what pace. The best way to get this information is by regularly speaking to the professor that teaches the lectures. Another reason to speak to the professor is for espionage. She can tell you what problems the students are interested in and you can let her know what concepts the students are having the most difficulty applying to the homework problems.

Prevent Cheating

The most important principle regarding cheating is to do everything you can to prevent it. If I catch a student cheating in my class (which happened a couple of times), I consider it a failure on my part (and the student's part, of course).

(To students: most cheating is much more obvious than you realize. When 50 students are writing a quiz, their heads tend to make similar movements when viewed from the front of a class. A student who is trying to read equations from a hidden sheet in their hand, or from an open textbook on the floor, etc., makes head movements that stand out. It is almost comical how obvious it is when students cheat.)

Run Quizzes With Stern Authority

I tried to be affable in all my interaction with students, but I found that the only effective method for running a quiz was to be stern and without humour. If you have a better way, please let me know. The goal in running a quiz is to make sure all students are given a fair and equal opportunity to show what they have learned. There are a few critical portions to running a quiz:

Allow 5-7 extra minutes above and beyond the length of the quiz for handing out and collecting the quiz.
Announce that the quiz will be handed out and there is no more communication allowed.
Hand out the quizzes face down.
Announce the length of time for the quiz and begin the quiz.
During the quiz, ask each student to present school identification and sign an attendance sheet by printing his name and signing. Check to make sure that the name on the sheet matches the name written on the quiz. This step has a few purposes. The identification ensures the student writes his own quiz (and not someone else's). Making them sign a sheet help ensure that they don't try to take the quiz with them and claim that you lost it.
Warn the students when there are 5 minutes left and that they should be prepared to stop writing when you announce the quiz is over.
Announce the quiz is over and make sure that every student stops writing. It's not fair to the class if a student keeps writing beyond the end of the quiz. If a student keeps writing despite your requests to stop, simply say “please see me at the end of the class,”, which usually causes the student to stop writing. After class, you can let the student know that if he repeats his behaviour on the next quiz, he will have to deal with the course instructor. If If the student still keeps writing, simply make note of it, inform the course instructor and let her deal with the situation.
Collect the quizzes carefully. Keep enforcing silence and make students stay in their seats until all quizzes are collected. This is the final opportunity for a student to pocket his quiz and claim that you lost it.
When all quizzes are collected, thank the students for their patience.

Avoid Email

Some students like to send email. They like it so much that the day before the quiz they write a 1000-word essay-style email regarding their thoughts on questions 25, 30, and 46 of chapter 5 of the course textbook. They expect an email response of similar length, that evening of course, because the quiz is approaching fast. You can imagine that answering even one email like this could take hours, let alone answering multiple emails. I dealt with these types of emails by inviting the student by email to bring his questions to the tutorial portion before the quiz. I took note of the problems of interest and made special preparation for them. This way the whole class benefits from your efforts, rather than only a single student.

Give Two Extra Tutorials

Since I tried to minimize my email and outside-of-tutorial interaction as I usually had plenty of other work to do, I had a policy of offering two extra tutorials during the term. One the day before the midterm exam and one the day before the final exam. About half of the class would attend these extra tutorials and seemed to find them useful to shore up gaps in their understanding after doing the pre-exam studying.

Mark Quizzes and Assignments Quickly and Fairly

You may find yourself in the situation where you have to come up with a marking scheme for a quiz or an assignment. Usually, the total number of marks is set, i.e., the quiz is out of 10, but you are responsible for deciding how to dole out the individual points. The best way to do this is to start with an arbitrary scheme, e.g., four marks for finding the output resistance, two marks for finding the gain, and four marks for finding the -3 dB frequency. Next, mark a few quizzes without actually writing anything on them, but recording the marks on a separate sheet, and see how it goes. I usually go through about 10 percent of the quizzes like this. After you've ghost-marked these quizzes, go back and re-evaluate your marking scheme. What's the average so far? Does the marking breakdown fairly reward the distribution of work? Does the marking breakdown reward the grasping of concepts rather than rote computation? Are at least two students going to pass the quiz? Now adjust the scheme to better meet your target average and to be consistent and fair. This pre-marking adjustment is much more elegant than the traditional after-the-fact bell curve. Finally, adjust the marking scheme so that you can complete the marking in the time allotted. Often, with only some small tweaks based on the quizzes you have seen, you can simplify the marking scheme to save you time.

As a courtesy to students, return quizzes in the next tutorial session after they are written.

Be prepared to re-mark some quizzes, since you will make mistakes. Some students consider the awarding of a single mark to be as important as life itself and are willing to sacrifice any amount of time to argue with you about it. I avoided this situation by adopting the policy that students had to hand in their quiz to be re-marked with a short explanation of why more marks are deserved. This diffused any confrontations as I could examine quizzes on my own and decide without pressure.

Ask For Feedback on Your Teaching

Ask the students for feedback about your teaching. One way to do this is to take five minutes at the end of a tutorial and ask students to anonymously write down what they like about your tutorials and what they want to see improved. Of all my suggestions in this guide, this is the only one that I didn't do and I think it was my biggest mistake. I would have been a better TA with the students' feedback.

Be Prepared

I told you being prepared was important, so to prepare you, I'm repeating this advice. Be prepared. Don't say I didn't warn you. Prepare!

Acknowledgments

Little of what I have written here is original. Most is begged, borrowed, or stolen from all the various instructors and TAs I have had the pleasure of working with. Specifically, I would like to thank for their knowledge Professors Ali Sheikholeslami, Glenn Gulak, Sorin Voinigescu, David Johns, and K.C Smith. Also, I thank fellow tutorial TAs Shahriar Mirabbasi (now Prof. Mirabbasi), Marcus van Ierssel, Yadi Eslami, and Trevis Chandler. Finally, thanks to all the unsuspecting undergraduate students on whom these techniques were honed.