People often ask me for recommendations on pedagogy. To avoid repeating myself, I’m going to make some suggestions here, and update this document over time.

1 Key Advice🔗

I think the single most thing important I can say is this: Every time you are inclined to use the word “teach”, replace it with “learn”. That is, instead of saying, “I teach”, say “They learn”. It’s very easy to determine what you teach; you can just fill slides with text and claim to have taught. Shift your focus to determining how you know whether they learned what you claim to have taught (or indeed anything at all!). That is much harder, but that is also the real objective of any educator.

2 Readings🔗

Over the years I’ve accumulated a small list of books that I like to recommend:

The ABCs of How We Learn by Daniel L. Schwartz, Jessica M. Tsang, Kristen P. Blair. The “ABC” is a gimmick and wears thin in predictable ways, but it also forces the authors to limit how many entries to have. Don’t let this get in the way of high-quality advice from experts.

How Learning Happens: Seminal Works in Educational Psychology and What They Mean in Practice by Paul A. Kirschner, Carl Hendrick. This book wears the research even more explicitly on its sleeve, but every paper is distilled into actionable content.

Understanding How we Learn: a Visual Guide by Yana Weinstein, Megan Sumeracki. A third way to organize similar ideas. Also combines cognitive science and education.

Each one has the following important charactersitics:

Its content is research-based, not just opinion.

It is written in an accessible style, suitable for non-experts.

It provides actionable advice.

It’s fairly small.

Simply implementing a handful of key ideas (that are new to you) from these books will make you a much more effective educator.

In addition, I’ve also found these very useful and thought-provoking, though they are not as immediately actionable:

How People Learn: Brain, Mind, Experience, and School from The National Academies Press. This is a significant summary of the research on learning.

The Teaching Gap by James W. Stigler and James Hiebert. An important book about what goes wrong specifically in American schools.

3 Neuromyths🔗

Education is replete with “neuromyths”. One of the most pervasive is that of “learning styles” (when I said “visuals” above, did you think, “Aha, that should appeal to visual learners!”?). Just type “learning styles debunked” into your favorite search engine and you’ll find dozens of articles saying so, yet the idea refuses to go away. For a very sober, cautious summary of the (lack of) research evidence for this idea, see this excellent review paper.

4 For Computer Scientists🔗

Specifically for computer scientists, I have some very specific things you should read to gain perspective:

Justin Reich’s Failure to Disrupt: Why Technology Alone Can’t Transform Education, which traces and shows the follies of a certain mindset that is pervasive in computing’s view of education. Odds are, even after you read this, you’ll still tell yourself, this time it will be different. In doing so, you will fail to understand that these are issues about humans, not about technology. Well, what can I say: I tried.

Morgan Ames’s The Charisma Machine: The Life, Death, and Legacy of One Laptop per Child is an excellent deconstruction of the OLPC project, which was an apotheosis of technological solutionism. Most of these issues could be, and were, predicted at the outset, but computing valorizes sunny optimism over knowledge and experience.

Morgan Ames also wrote Hackers, Computers, and Cooperation: A Critical History of Logo and Constructionist Learning. This deconstructs a pervasive mindset in computing education for its strengths and (many) weaknesses. If you don’t have the energy to read Ames’s book, at least read this article (because OLPC is the product of this mindset combined with lots of resources).

Audrey Watters’s The 100 Worst Ed-Tech Debacles of the Decade. It’s easy to imagine these as ZIRP phenomena, and many are, but (a) the underlying mindset is pervasive in computing and (b) every decade spawns a new set of technologies and views that can spawn a new one of these articles.

5 Classroom Tips🔗

Over the years I’ve adopted several classroom techniques that appear to have been useful to my students. In what follows I will refer to the educator as “you”, on the assumption you’re reading this in the hope of getting new ideas. Anyone is welcome to read this, but the person with the power to make a change is you!

To start out, I’ll make a provocative statement. Most educators think a great idea is to ask a question and get an answer or two, and when asked to teach “actively”, this is what they do. I don’t want to discourage you from doing this; if this seems innovative to you, then it’s probably better than what you were doing before, and you should do it! But I think you can do better. Read on.

Problem:

Coming to class and just staring at you without any idea how to answer your questions is a discouraging experience. Also, it creates a power dynamic of you always being in charge.

Solution:

Remove the isolation factor. Also, reduce the amount of time you’re in charge.

I use a variation of Think-Pair-Share. Ask a question, and give them time to talk it over. This has several unexpected benefits:

Whether or not you (can) walk around the room, you get some sense of what students are saying. This will help you realize if they have some misconceptions, but even more importantly, it will help you realize if you didn’t make the question clear enough! If a lot of them are talking about the “wrong thing”…it’s probably not them, and you should bring the class back together to clarify what you want them to be talking about.

Just by listening to the conversation, you can tell when they’ve had enough time. Once they’re done (enough), the conversation will often drift off-topic. You don’t have (and can’t afford) to let every conversation get to completion. Keep listening for the temperature of the room, and cut off at some reasonable point. You don’t want to wait until they’ve exhausted the topic; you want to cut off just past the peak.

You should try to have some interaction every ten minutes or so, at the latest. Usually when I violate this guideline I can feel by at the end of class: that something wasn’t quite right.

Problem:

Class isn’t only about learning and answering. It’s also about socialization. Working with partners can be awkward.

Solution:

To make the social aspect work, start every class by asking students to introduce themselves to their neighbor. I really mean introduce: say hi, tell them where they’re from, that sort of thing. Really do give them a minute to say something about themselves as humans.

Do this at the start of every class for a few days. Let them get to know one another and initiate friendships. They’re in your class because they have something in common; let them discover what else they have in common.

Even after a few days, before asking your first question, ask students to tell their partner their name. Why do it even if they’re sitting by the same person as yesterday? Because it’s embarrassing to admit that you’ve forgotten the other person’s name! Being told to introduce themselves again absolves them of that tension at the very outset. They’re not introducing themselves because they’ve forgotten their neighbor’s name; they’re doing it because you told them to!

Problem:

That one student who dominates q&a. Note that just getting the room as a whole to discuss is already an improvement, because everyone has had a chance to participate. But there is also a bit of tension in a classroom when you ask a question: who’s going to relieve the class of the pressure? If there’s a “safe hand” at the front, then other students no longer feel the need to participate, and indeed, over time, don’t.

Solution:

Limit the number of times a student can answer a question on a given day.

I derived this idea from Federico Ardila-Mantilla via the article Math Is Personal (I recommend the whole thing). My rule is that when you raise your hand to answer a question, you have to raise as many fingers as the number of questions you have already answered that class.You can ask as many questions as you want, though! No fingers for those: they don’t count. That is, the first time a student raises their hand on a given day, it looks like a fist.

My rule is that I prioritize the lower number of fingers. But I also say that raising one finger is fine, but raising two is really pushing it: it means you want to give an answer for the third time that class. In that case, I say, I really will be judging you; you better have something really interesting to say. (But that’s also why you can: maybe you do have something really interesting to say, and we’d all benefit from hearing it! It doesn’t even have to be right; an interesting wrong response is also very educational for everyone.)

This also creates some vocabulary for the class. “Can I see more fists?”, “Hmm, I see too many ones.”, etc. all become natural things to say and are automatically understood. But it also forces the power to be spread across more students. That small group of students can no longer hog the class.

Problem:

Class becomes a game of “Guess what I’m thinking of”; guess correctly and you get rewarded, guess incorrectly and you openly get told that. That sounds horrifying, right? But you probably do that without realizing it.

See, this is the difficulty with asking a question and taking answers until you get the right one. Very quickly, students figure out the game. This then has two adverse consequences: students who aren’t confident will never speak up, and those who know the answer (or think they do) will jump to tell you, and the whole classroom atmosphere goes south.

Solution:

Don’t take just one answer; take all the answers. Take answers until you’ve gone on too long or the hands have stopped being raised.

Try to stay as unreactive as possible when you get an answer. Treat them as interesting, not as correct or wrong. The moment you start passing judgment, you’ve reverted to the game. (The two exceptions I make are: (a) if the answer is completely off, I might nudge the class back on track; (b) if the answer is especially interesting, like something I have rarely heard someone point out, only then will I give the student a bit of a verbal pat on the back.)

Instead, take all the answers and “summarize” them back to the class. Usually, I find that I got a bunch of partially correct answers, some totally correct ones, and some wrong ones. But by putting them all into the cement mixer of my summary, they lose their individual nature. Each student can decide for themselves how correct they were, and the class still gets the right answer in my words, which have probably been thought through more carefully than most student answers.

An added advantage to this mechanism is that the rest of the class won’t know the right answer until the end. Therefore, they need to stay engaged and reflect on every answer as it comes in, because I’m not doing it for them. This may also inspire new answers! That’s the best case, because it suggests students are actively formulating and testing hypotheses and understanding.

Problem:

It’s useful to get the temperature of the room periodically. It also healthy to force them to commit to an answer, even if it’s the wrong answer: I learned this from Mark Guzdial, who was channeling Eric Mazur.

Solution:

Easy! Get them to raise hands!

However, this has two problems. First, everyone can see all the people around and in front of them, and will know if they voted incorrectly. That will heighten any self-consciousness or lack of belonging that students feel. So you want them to be able to vote without revealing too much.

Okay, use a clicker or some other electronic device! These days there are clicker apps that run on your phone or laptop, even!

Reader, you have so lost the plot here. Electronic devices are pits of distraction. You want to do everything you can in class to keep your students from returning to them. I feel certain that the worst thing you can do is to have them open their devices on your order.

And here’s the second problem with either raise-hands or electronic voting: it isn’t social, it isn’t fun.

No, instead! I got this idea from Biella Coleman, who pointed out that this is how the IETF works. You vote by humming.

This addresses all the above issues, and does more:

No electronics involved.

You can vote fairly confidentially.

It’s fun! People want to make noise! It keeps them awake, and it makes them feel alive.

Very subtly: The volume of a hum is a continuous variable. If a student feels particularly strongly about something, they can hum loudly. It’ll happen, and when it does, if you treat it in good humor (as you should), everyone will get a good laugh out of it, and it’ll encourage a certain kind of great lawlessness.

See, the discrete counts never really mattered that much. Don’t fall for the trap of false precision. It doesn’t matter whether the vote was 32 for and 31 against or vice versa. Humming keeps the internet going, because the people who invented it had some deep intuitions about humans and decision-making. Take advantage of their intuitions.

Problem:

You want to know whether people understand a thing. You ask “does everyone understand?” A fraction of hands go up; the others don’t. Now what?

The difficulty here is that you’ve become like a security alert. You’re asking a useless, almost impossible, question. It’s blocking progress. It may cover your ass, but it doesn’t produce useful, actionable information.

Solution:

If you must ask this question, at least start by asking the negative question. If you ask “does everyone understand”, that’s an absurd question: nobody is in a position to say whether everyone understands, only themselves! If you ask “do you understand”, there are all sorts of reasons why people might not raise their hands; you’re left in an awkward position of not knowing what to do. But if you ask “does anyone not understand”, even one hand is useful feedback for you.

However, it’s much better to not frame it so starkly. First, don’t ask about “understanding” in general; be specific. (A student who thinks they understand may not even realize what you were asking about! You know what all the pieces are; they don’t.) Second, give them a range of possibliities; “understanding” is not a binary activity. Third, use techniques like hum-voting so that people don’t feel singled out.

Indeed, if you do periodic hum-votes, you won’t need this very awkward question (in any form) at all. Think of student understanding as a potential random walk. However, your class plan isn’t that at all: you have a very specific place you’re trying to get to and a small set of routes you think is the best way to get there. Therefore, at every point you want to try to consolidate the walk by curbing the randomness. Asking intermediate questions and getting feedback helps you correct confusions before they compound. And then you never need to ask the big, global question.

Problem:

All this sounds like it’ll take an awful lot of time!

Solution:

Yes, it does. 1.

But you can still get through an awful lot of material. I do. 1.

But most of all, please re-read the Key Advice! These are the tools I use to achieve that goal. What’s your real goal?