The Problem With Lectures

Yesterday’s physics education post kicked off a bit of discussion in a place I can’t link to about the usefulness of lectures. Something in that reminded me of an anecdote from my grad school days, that I think is useful, so I’ll post it here.

When we were working on the spin-polarized collision experiment, we expected that a certain effect ought to be pretty big, but when we did the experiment, it didn’t show up at all. We mentioned this in a talk or poster, and a theorist said “Oh, that’s not that surprising.”

We had been pretty surprised, so I was sent off to talk to Paul Julienne, one of the outstanding local theorists, to get an explanation. I went up to Paul’s office, and spent half an hour talking to him about it, and the explanation he gave was perfectly clear, and made total sense.

So, I went back downstairs to Steve Rolston, my immediate supervisor at NIST, and repeated the explanation to him. And somehow, it didn’t make a lick of sense when I said it. Steve poked holes in it immediately, and after five or ten minutes, I didn’t agree with what I had said any more.

After that, Steve went upstairs to talk to Paul, and after half an hour came back downstairs saying he understood it completely. When he repeated the explanation to me, though, it didn’t make any more sense than when I repeated it to him.

We went through a few rounds of this, and never really did get it down until we made Paul come downstairs and spend an hour at the blackboard in the seminar room, where we all hashed it out together. After that, we knew what was going on.

What’s this have to do with lectures and my students’ complaints? Well, far too often, lectures and recitation sessions are just like the conversations Steve and I had with Paul. When somebody else is presenting a detailed explanation of how they solve some problem, it’s very easy to nod along and say “Yes, yes, of course, that’s the thing to do.” You leave the room perfectly convinced that you’ve understood everything, but when you try to apply what you think you know by explaining it to someone else, you find that you didn’t really understand a thing.

(Another example, from my angry QM prof when somebody suggested The Feynman Lectures as a text: “The thing about Feynman is that you read it, and you say ‘Yes! I understand! I’m doing physics!’ then you try to do a problem and you find that, well, you’re not Feynman.”)

My statements about the limited utility of problem-solving in lecture were not intended as a knock on my students, which is why I think this is useful. At the time of this anecdote, I was this close to being a professional scientist– in my fifth year of graduate school in physics. And I still fell into the trap of thinking I understood something just from a verbal explanation.

That’s the problem with good explanations: they’re incredibly seductive, convincing you that you understand things that you don’t understand at all. That’s the logic underlying the active learning methods– making students explain things to each other in class discussions leads to a deeper understanding than just listening to a professor explain the right answer. And that’s why going over problems at the blackboard is so attractive to students– they leave the room thinking they understand how to do the problem– without actually being any help at all.


  1. #1 Dan
    November 30, 2011

    But you had to have a long lecture from Paul before you understood it. If your premise that active learning is better for students, then you and Steve would have been able to work through the problem with greater understanding.

    Lectures are useful not because they promote significant insight in the students, but because they convey a lot of information in a short time. No one who uses lectures to teach, expects students to get it the first time. That is why homework and term projects are assigned.

    As I see it, the problem with active learning methods (as a sole teaching technique) is that they never provide the students the insight that the teacher has. This is a key part of taking any course. As a teacher explains things, he or she gives transmits tacit knowledge that would take students a lifetime (if ever) to acquire. There are ways of viewing the world that make approaching problems easier, and a good teacher helps convey that to the student so that when they approach a problem they have a line of attack, even if they don’t necessarily understand every thing first.

  2. #2 Chris Goedde
    November 30, 2011

    Dan, what makes you think that active teaching methods means that the instructor never provides insight or explains things? I also think you’re missing the point of Chad’s story: when Paul went downstairs, they “hashed it out together” on the blackboard. That doesn’t sound like a lecture to me.

    I don’t lecture any more, but I still talk a lot in class. (Well, sometimes I give mini-lectures on certain topics.) Sometimes I talk to the whole class, sometimes to individuals or small groups. I certainly hope that I share my insights and explain things.

  3. #3 Chad Orzel
    November 30, 2011

    Lectures are useful not because they promote significant insight in the students, but because they convey a lot of information in a short time.

    The question, though, is whether that represents a real and useful sort of efficiency, or whether that’s just an illusion. That is, it may seem like you’re getting through a lot more material, but if very little of it actually “sticks,” are you really doing any better than you would with a slower method that produced more retention.

    There’s a fairly convincing (to me, anyway) argument to be made that a lot of lecture time is effectively wasted, with all the real learning coming after class, when students try to solve problems by piecing together material from their notes and the textbook. The goal in more active approaches is to have them get the basic stuff from the book on their own, spend class time working through the trickier points, and thus make more effective use of the class time.

    I would also take issue with the idea that there’s no instructor input in the active methods. When things are working well, the instructor does a lot to shape the discussion even if they’re not directly telling students “Do it like this.”

  4. #4 Natalie
    November 30, 2011

    But what do you do in a subject which isn’t problem based? I teach biology, and I can’t wrap my head around how to apply problem based teaching to a class which is mostly about getting students to learn the facts.

  5. #5 Chad Orzel
    November 30, 2011

    But what do you do in a subject which isn’t problem based? I teach biology, and I can’t wrap my head around how to apply problem based teaching to a class which is mostly about getting students to learn the facts.

    I can’t speak to biology teaching directly, but there is material on educational reform from the Life Sciences departments on the Colorado education site, which might be useful.

  6. #6 Sue VanHattum
    November 30, 2011

    Natalie, there are lots of misconceptions about biological systems. (You probably know that better than I do.) I think you can judge your success on how well your students overcome those misconceptions.

    I’m a math prof, not a scientist, and I would have told you (until about a year ago) that trees got their mass from what they pull up out of the ground. I’d have been wrong…

    I love the blog Science Teacher. It’s about high school science, but I’ve learned a bunch from it.

  7. #7 Chad Orzel
    November 30, 2011

    Sorry*, that’s the University of British Columbia, not Colorado. I got there via the Colorado physics education site, but it’s not their site.

    (*- pronounced “sore-y,” because it’s an apology to Canadians…)

  8. #8 JM
    November 30, 2011

    Chad- As you’re a Williams alumus, is what you described an example of the benefits of the physics tutorials at Williams?

    That is, is this the scenario– Topic for the week, overview or lecture of topic presented by professor, students given problem set (and from what I gather a very difficult set of problems), then a few days later pairs of students meet with the professor after they’ve done the set and they all work together to grapple with the topic via the problem set.

  9. #9 jim
    November 30, 2011

    After I’d been teaching for a few years I decided that lecturing was probably the least useful thing that I did (coming up with good assignments and providing feedback on grading them was probably the most; office hours, too, for those students that actually came to them), yet I still spent most time on preparing lectures, because that was where my ego was exposed. I never figured out how to get round this.

  10. #10 Kaleberg
    November 30, 2011


    I always found lectures extremely useful both in understanding and problem solving as did most of my peers. Many lectures were extremely focused on the mechanics of attacking and solving problems, particularly those in physics and signal processing. I remember one charming lecture on how to deal with Fourier analysis geometrically, and I can still explain how a crystal set (or super-heterodyne or single sideband) works after 40 years.

    My current experience with recent vintage students is that they are not very good at using text books. I’m often translating their own textbook or class notes for them as if from a foreign language. Are they having the same problem absorbing the spoken word as well? Listening for information is a skill, just as reading for information is a skill, and like most skills one can get better at it with practice. While I would consider active participation a form of hell, it is possible that it might help some students develop the listening skills they need.

  11. #11 Cherish
    December 1, 2011

    I’ve run into a couple studies in the past year or two talking about how students tend to think that because they understand a concept after seeing it presented, they can easily use and/or apply that concept in an exam. Few people ever learn to use retrieval and self-evaluation in their study skills, and few students ever are presented with the idea of metacognition.

    As teachers, we’re sort of allowing the illusion to exist that because if they understand a lecture, they must understand the material as well…otherwise, why would we lecture? But honestly, I think very few students have ever been told point blank that working things out on their own (sometimes repeatedly) is really the key to good understanding. And most high schools don’t teach this skill, so kids are very much taught that the key to success is to sit back and listen while being spoon-fed knowledge.

  12. #12 becca
    December 1, 2011

    I don’t know that the conceptual understanding gained from lectures is illusionary, so much as different from the procedural understanding gained from practice sets.
    It’s just that most exams, composed mostly of problems, require you to have the procedural knowledge down pat and a bit of conceptual knowledge. It’s just as possible to ask for short essay questions describing situations that would emphasize conceptual knowledge and minimize procedural knowledge.
    I suspect that, the ability to listen to a lecture, and ask a question about “what would happen if…?” and be told that it’s 1) currently being pursued by the lab or 2) a great idea is a valuable ability for a working scientist. Conceptual knowledge helps you ask questions, and even identify appropriate tools to solve them- if the tools to solve them exist (even if you don’t necessarily know the procedure to apply them yet).

    That said, it’s my private observation that listening attentively to lectures is both very engaging and very intellectually tiring. I think part of the uselessness of lectures may be that we do try for such high throughput.

  13. #13 Dan
    December 1, 2011

    I also think you’re missing the point of Chad’s story: when Paul went downstairs, they “hashed it out together” on the blackboard. That doesn’t sound like a lecture to me.

    Chris: No. I don’t think I did miss the point. And, yes it does sound like a lecture. Effective lectures do have interaction between the lecturer and the class. The lecturer, if they are doing their job, is engaging the class getting them to think about a problem in ways they haven’t before. One advantage, however, is that you can reach a lot of students at once. While this may not be one-on-few like the example of this blog, lecturing can be a very effective way to convey information. To bad you don’t lecture anymore. You are missing an effective tool for education.

    The question, though, is whether that represents a real and useful sort of efficiency, or whether that’s just an illusion.

    Chad: Yes, that is the question. To say that all lectures represent the illusion is illusory. Lectures can be very efficient use of class time; so can active techniques. However, they can both be very inefficient. The educator needs to know what technique achieves the desired goal. My point is that lecturing is not a priori bad.

  14. #14 Mark
    December 1, 2011

    This reminds me of a possibly apocryphal story one of my profs told me of a grad student in Germany (I think) who complained to his advisor that he didn’t understand some particular area. His advisor said something like, “Excellent! Then you shall teach that subject next time.” There’s no better way to learn something, or to know that you haven’t learned it, than to explain it to someone else.

  15. #15 CCPhysicist
    December 3, 2011

    I just don’t get the point of “learning facts” (mentioned @4) in lecture. What is the difference between watching PowerPoint live and reading a book or watching a video? Sometimes I think I know the answer, but all it takes is a quick quiz to find out that I am wrong. They learned zip.

    However, your example is just as much an illustration of the failure of active learning as it is of the failure of lecture. What it really shows is that you don’t get it the first time, and it might take a week before it sinks in.

    I think some of the problems we face go a lot deeper. I am a firm believer that students need to see a model solution, but I also know that they don’t believe what we do is how problems get solved. (They think we know 1000 magic equations that answer every possible problem, and we just do all that problem-solving stuff for show.) You see, they have been using math books for a decade where each problem has a pointer to show which example to follow to “solve” it. This is now evolving into on-line systems that will show you a “similar problem” — that is, an example identical to the problem in question so you never have to actually SOLVE a problem when doing the homework! That is what they expect in physics and engineering, a detailed template for every possible problem, as if every bridge appears as an answer in the back of the book. They don’t realize that active learning is what we do every day.

  16. #16 Ovgu_victor
    December 7, 2011

    Part of the uselessness of lectures (esp in germany) is that you have to copy the notes from the blackboard al the time, thereby preventing you from paying attention.