At M.I.T., Large Lectures Are Going the Way of the Blackboard | Hacker News Reader

At M.I.T., Large Lectures Are Going the Way of the Blackboard

nytimes.com

46 pointstokenadult46 comments

The general sentiment about the TEAL classes (the ones with the clickers) is that they are really gimmicky. Supposedly, it gives the professor some feedback on what concepts are unclear and pushes the students to be attentive.

The problem seems to be that it's just unnecessary structure for something that's typically dynamic. It at least partially devolved into an attendance system - students used to just press in for missing friends, until recently when they had each student buy their own clicker.

I remain convinced that the only reason they require those clickers is to charge $70-$100 to each student for them.

You're joking! $70-100?

As a slight thread hijack, which do you prefer, blackboards or whiteboards?

The physicists and mathematicians I know seem to universally prefer blackboards, while everyone else seems to like whiteboards.

I like whiteboards for any semi-permanent or "data-rich" writing (i.e. multi-color), but probably prefer blackboards for intense computation.

While RK is asking - has anyone had contact with those glass boards that Joel wrote about recently?

Yep, they're nice.

One thing though, when there's direct light, what's written on the surface casts a shadow on the painted back-side. It looks a bit odd.

Also, the same effect results in a fuzzy picture when the glass board is used as a projector screen.

I'd love to experiment with back-lighting the glass board. If would eliminate the shadow-problem, and look pretty neat, I think.

We've had them in my previous company. How different are they from the regular whiteboards?

They probably look a lot cleaner than the white ones after years of usage. If the ink in the marker dries out, it is harder to read than it is to, from a whiteboard. Can't think of others for now.

I like blackboards --- but then I am a mathematician. I even brought blackboard foil to glue to the walls at home.

i like blackboard for the simple reason that chalk never lies

blackboards are much better:

* easier to erase

* easier to clean board and erasers

* chalk is cleaner, cheaper, more expressive, less wasteful (no plastic), and there are generally more colors available

* chalk stays at full "marking capacity" until it runs out. Dry-erase markers make you write at less than full contrast for a long while until you toss them.

* chalk dust is no problem if you have windows you can open a little

* sounds better when you use it. :)

* nicer light-on-dark contrast (IMO)

But whiteboard gets chosen by non-experts b/c it supposedly makes less dust for electronics/projectors. Unfortunate.

>But whiteboard gets chosen by non-experts

Presumably, non-experts in the details of your opinion about whiteboards vs. blackboards?

"the professor covered multiple blackboards with mathematical formulas and explained the principles of Newtonian mechanics and electromagnetism."

Walter Lewin did a lot more than just that.

Walter Lewin did a lot more than just that.

Please tell us more. I've never had the privilege of seeing Walter Lewin in action.

Professor Lewin has several classes available online through both MIT OpenCourseWare (OCW) and the iTunes store.

OCW Link: http://ocw.mit.edu/OcwWeb/Physics/8-01Physics-IFall1999/Vide...

As entertaining as his lectures were, it was still pretty easy to leave with barely any more knowledge than you had going in. Especially as the topics got more complex.

I had him for my Waves class, and kids found it to be incredibly hard (and to require lots of on-your-own learning) even while Lewin was jumping around the room and making cool things happen with light and sound.

The large lecture hall is an example of something which was necessary due to past technology constraints but which persists even though it's no longer optimal. I've taught such classes of 75 to 80 students. Even at that modest size, there is considerable overhead just from managing that many people.

What we really need are uses of technology that maximize interaction between students and professors and amongst fellow students. The most productive time in terms of my teaching were the lab sessions of only 25 students each. That and office hours, which were attended by a shockingly small number of students. I suspect that study groups are also highly productive. Technology should be used to enable more interaction akin to study groups and office hours.

"As each semester progressed, attendance in [the professor's] introductory physics courses fell to 50 percent, as it did, he said, for nearly all of his colleagues...'I had poor attendance, and was failing 10 to 15 percent'"

So, 50 percent of the students aren't coming to class, and they're only failing 10-15 percent? It sounds like the tests are too easy. If the tests were more demanding, more students would come to class--which might (paradoxically) lower the fail rate.

"Unlike in the lectures, attendance counts toward the final grade...[the professor] gauged the level of understanding in the room by throwing out a series of multiple-choice questions. The students “voted” with their wireless “personal response clickers”"

I'm currently a sophomore in Computer Engineering, and I have had several 150+ student lectures that used these clickers to take attendance. It's a simple way to make sure students come to class. This is a major benefit that has nothing to do with smaller class sizes.

It sounds like the tests are too easy.

You might be right that much or most of the higher pass rate is due to higher attendance, but it's also possible that students were learning plenty from section, the book, and working on problem sets. When I taught the analogous course (Physics 1) at Caltech, many of the students skipped lecture entirely. The failure rate was typically under 5%, and I guarantee it wasn't because the tests were easy.

One student I remember took it all the way: he skipped everything, including homework (which counted for 20% of the grade), but he got perfect scores on all four quizzes and on the final exam. According to the numbers he had an A- because of the 0 on homework, but everyone in the grading room agreed that was absurd, and so we gave him a flat A instead.

1. - 0.2 = 0.8 is typically a B-. Is there a different grading scale there?

curve?

Basically all classes at Caltech are curved. It actually makes a lot of sense, because it means you can make the tests really hard and leave plenty of room at the top, which among other things helps reduce the variance due to careless errors. (If you make a couple minor minus sign mistakes, but nail a hard problem most people miss, you still come out ahead.) In Physics 1, typically 90%+ was A+, ~82-89% was A.

I pulled this trick in my entry-level calculus class in college. They were so generous with extra credit that my marks should have meant that I got an A. But I didn't do any of the homework and the TA decided to punish me with an A-. I suppose I should have fought, but I'm epically lazy.

Lesson: if you're going to do something like this, don't assume that the people on the other end will hold up their end of the implied bargain. Either fight or do some token tribute.

or even their end of the /explicit/ bargain. I had a prof fail me, after announcing that the final grade would be the class grade. I never went to class, so he felt justified in ignoring my A-/B+... took me awhile to get over that slight.

The intro physics class at Harvard is graded something like:

k * homework_score + (1 - k * homework_score) * (final_score^extra_credit)

So you can do no homework and no extra credit as long as you ace the final.

It sure wasn't that way when I was there. ;-)

First year physics is mostly review for a lot of incoming freshmen, especially if they did one or more AP classes in high school. I'm not suprised that some would do great even skipping the lectures.

When did they put freshmen on grades? When I was at Caltech years ago, first year was pass/fail.

Back then, Goodstein taught phys 1a. Fine lectures, unless you had read the chapter in advance. Then you noticed that his lecture was almost identical to the text, which made it hard to stay awake. Oh, and the retch session were taught by full profs, not grad students, so a lot of learning happened outside of lecture.

Highly qualified students have the option to test out of the different terms of Physics 1. (I know because I graded the placement tests for five years.) I suppose some of them take the course anyway to get an easy A, but this guy I'm talking about aced Physics 1c analytical track, an E&M course that uses Purcell, and very few high schools cover material that advanced. (It's well beyond even AP Physics C material.)

Regarding grades, here's how it shakes out nowadays:

  1st term: pass/fail
  2nd term: "shadow" grades
  3rd term: real grades

Shadow grades don't appear on your report card, and they don't count towards your GPA, but they are used for internal scholarship awards and recommendations.

Recitation section TAs in physics have been a mix of professors and grad students for as long as anyone can remember. Virtually without exception the best TAs were grad students (as measured by student reviews, section sizes, and exam results), though some of the professors weren't bad either. This jibed with my undergrad experience at Harvard: almost all my physics professors sucked, and by far my best instructor was the 22-year-old grad student who taught my freshman physics section.

While I can see where you're coming from, a few other factors need to be taken into consideration.

One of MIT's greatest strengths (and arguably weaknesses) is the sheer "atmosphere" of working on your classes. The line between directly school-related stuff and other stuff is very blurred. While of course it's true that homework and tests impact the lives of students at any and every school, I think it's fair to say MIT takes it to another level. And not just with direct things like giving lots of homework and tests - I often felt like it was easier to find an Athena cluster (what MIT computer labs are called) than a drinking fountain or a snack machine. Because of this atmosphere, the emphasis on necessarily attending lecture is lessened, as there is more of a "continuous" feel to each class, which might include going to lecture, recitation (which is often much more valuable than lecture but doesn't get included in the "50 percent" quote from the article), study groups, office hours, and more. Now, does this devolve into problem set copying for some students? Definitely. But overall the atmosphere was definitely there. Because of this, it's misleading to focus on a number like "50% of students aren't coming to class", as attending lecture is just one part of the equation.

I should add as a disclaimer that I never took a TEAL class at MIT, or even a physics class (something I regret). So, while I have perspective on the school as a whole, my perspective is still that of an outsider when it comes to the physics department.

As a side note, personally, I kind of hated the atmosphere because people often adopted a "woe is me" attitude as semesters went on. You'd be amazed how hard it is to get an MIT student to commit to spending an hour or two attending a potentially worthwhile event, in spite of the fact that they'll easily burn that much time complaining about their workload in just about any given week. At the same time, I liked the overall philosophy which really seemed to encourage pushing yourself to the limit. The Institute allows things a lot of other schools don't, such as for a student to take 6 or more classes in a term without special petition or paying extra, as well as for students to take classes that conflict (it's pretty hard to fit 7 or 8 classes into a schedule otherwise)....

Everyone seems to be forgetting that students that stop attending usually drop the class -- so that they won't fail.

I got an A+ in one of my courses without attending a single lecture (it was in an inconvenient time slot... which is why I didn't even go to the first lecture), and I enjoyed the course a lot more than most of my friends who did go to the lectures.

Don't overestimate the value of lectures. Sometimes a bad professor can do more harm than good to the learning process, especially for something you actually find interesting.

And, in a virtual age, what is the point of a physical school?

Answer: there is no point.

Synergy. Relationship. Feedback. Allowing smart people to meet each other today so that they will build the startups of tomorrow.

All possible online. No? For example, a lot of HN visitors interact here primarily to meet founding partners.

Possible but not as good as when you interact directly with people (in my view).

Direct feedback with teachers is far better than a lot of email exchanges, or than videoconference.

Spending a lot of time with other students on the same projects, in the same room and everyday at the university give you the opportunity to know their personality, and how they work. A lot better than doing a project in common with Internet.

You also have more opportunities to meet people from other horizons (i.e. non geeks).

Direct feedback with teachers is far better than [...] videoconference.

Wouldn't that depend upon the quality of the "videoconference"?

http://www.google.com/search?q=telepresence+%22eye+contact%2...

in the same room

Kind of like ... HP's Halo?

http://www.hp.com/halo/introducing.html

Halo gives the sense of being in the same room together.

Nerd sex.

There can be great value in personal interaction. I agree that big lectures like described should and will die off. However, the fastest way to learn anything is to have a series of conversations with an expert. (I guess that could still be reproduced by videochat or something...) I understand this has long been the model in many UK schools.

All that said, in a virtual age it is perfectly possible for someone to learn with out any physical school or personal interaction, just slower.

<i>There can be great value in personal interaction. I agree that big lectures like described should and will die off. However, the fastest way to learn anything is to have a series of conversations with an expert. (I guess that could still be reproduced by videochat or something...) I understand this has long been the model in many UK schools.</i>

The long conversations with an expert are purely an Oxbridge thing, it costs waaay too much to do that for it to be worthwhile for students with less potential/no tradition of doing so.

http://news.ycombinator.com/formatdoc

"(I guess that could still be reproduced by videochat or something...)"

Even if it is delivered as videochat, the attention of the instructor is still a limited resource. And this is the value proposition of a 21st century university. I think this change to the MIT lecture format reflects an understanding of this. Now that MIT has made much of its content available as Podcasts, etc., they are looking for ways they can provide more value to the people on campus.

Absolutely, and that is great for students. It could be that universities will become more relevant in a digital age. In the past, a student sitting in a gigantic lecture hall didn't have too much advantage over one just reading a book (and not paying tuition). If universities move to giving student 1 on 1 (or N on 1) interaction, someone in a university will have a big advantage over someone outside.

I've been reading Stanford's Engineering Everywhere lectures, and the thing I'm most impressed about is how much support there is for students:

Course-specific computer labs with course-specific lab assistants to help you walk you through problems as you're coming up against them in learning to code!

Sections with section leaders who grade your code in front of you and discuss the strengths and weaknesses of your coding style!

There is no advantage to a huge lecture hall over a podcast. But in-person mentoring, one-on-one human interaction is the best way to learn if you can afford it.

in-person mentoring, one-on-one human interaction is the best way to learn if you can afford it.

Is it possible to tell the difference between in-person presence and immersive telepresence?

Is it possible to tell the difference between in-person presence and immersive telepresence?

Yes. I say this as a parent of a pioneer class student in Stanford's EPGY Online High School.

I'll acknowledge that the several examples I have seen (not just EPGY OHS) of distance learning courses have far from immersive telepresence, but I think there is evolutionary adaptation to real-world, in-person conversation that isn't taken advantage of by any distance learning communication mode I have seen.

Agreed. I'm a TA for a CS course here and we have open office hours where the students can come and do their homework and ask questions as they come up, which I can then answer on the board (I'm not going to say if it's black or white so as not to alienate those who apparently feel strongly about this issue), benefitting not only the student with the question but also whoever else is there. I've had students help each other in my office hours too. That's not something you're going to get with SEE.

Answer: It's far easier to talk to someone over a beer when you're in the same building as them. Or to pop up at their office unplanned with a question. Or to work on a project with team members.

School isn't all about lectures, and even for lectures, pre-recorded lectures don't let you ask questions.

Reply on news.ycombinator.com