UW-LaCrosse started it, launching an online math course, “College Readiness.” with the help of a grant from the Gates Foundation. A lot of the energy around MOOCs has centered on advanced courses: machine learning, business analytics, and so on. The kind of thing that gets funders and nerds excited. But funders and nerds are already educated! If MOOCs are to provide the educational equity they promise, they’ve got to do it at the low end — giving kids access to a better math course than their understaffed, underresourced high school can provide. Is there a demand for basic, unflashy math instruction online? Seems like it: a thousand people have signed up, twice as many as UWLC expected, including an 83-year-old and an 11-year-old. Those are much smaller numbers than Coursera gets for its sexy machine learning course, but I’ll bet the gap in number of finishers will be much narrower; this course is an “I need to,” not an “It would be cool to.”

And now UW-Madison has gotten into the act, announcing yesterday that UW would be offering four courses with Coursera, two to start this fall.

Readers: would it be a good idea, a bad idea, or some combination of both, for me to propose to teach a number theory MOOC?

### Like this:

Like Loading...

*Related*

Would you be compensated for it? Would it count as one of your courses for the semester?

I don’t know, and it’s obviously relevant, so feel free to answer conditionally.

I’d be interested in signing up. So there’s that.

By coicidence, our university just announced some offerings on Coursera as well, and so I’ve been wondering about this (with the appropriate substitutions in place of number theory). It seems like it would be a great thing to have done and good PR for all involved, but you might do nothing else for several months.

I would be _really_ curious about a detailed story of how implementing one of these went.

Ronen Plesser across the hall from me gave a MOOC on astronomy last term:

http://sites.duke.edu/introastro/author/plesserduke-edu/

It sucked up all his time for a few months, so be prepared for that. I mean, you’re a pretty energetic guy, but so is he. He did seem to enjoy it though. Also I gather it’s crucial that the A/V people you work with be really good.

[…] Wisconsin puts the MOO in MOOC (quomodocumque.wordpress.com) […]

I think that Dick Gross has an elementary number theory course on EdX.

Maybe I’m thinking of http://www.extension.harvard.edu/open-learning-initiative/abstract-algebra

Another question: do you have a sense of how the course could be robo-graded? I do not recommend doing the MOOC if you have to grade 50,000 assignments by hand each week.

No direct experience with MOOCs, but this the internet so I get to comment anyway. Personally, I’d be intimidated by the amount of work involved — I suspect that it’s probably closer to writing a textbook than just teaching a class. Do I notice that almost all the instructors on Coursera aren’t tenure-line faculty but hold various types of more teaching-oriented positions.

A friend of a friend (also a math prof) took one of the general-interest math courses on Coursera and came away pretty unimpressed with the basic setup (for example peer grading was apparently disaster). I think if you’re seriously considering this then actually taking a MOOC as a student is the place to start; Coursera has some interesting statistics and CS courses coming up, though perhaps it makes more sense to do one where one already knows the material.

I admit I’ve toyed with the idea of doing one of these too, even though I’m pretty skeptical of the whole MOOC concept. I already teach calculus to 500 students a semester, what’s another couple of orders of magnitude between friends?

I think that Nathan’s recommendation to personally take MOOC course is an excellent one. You would have a better idea of how it works on the student’s end. Without robo-grading, scaling up from 500 students to 50,000 students would require 100 times more teaching assistents for grading exams and problem sets, but administration would not likely approve that idea and insist instead that the teaching assistents work a little harder and spend 100 times more time on that task. After all, they’re paid handsomely and they have nothing else important to do anyway, right?

I was a math teaching assistent a long time ago. I think that many of the anxious students would have preferred multiple choice exams and those certainly can easily be robo-graded. However, multiple choice exams in math does nothing to encourage real effort by marginal students. Many of them would bet that they could at least get a D by random chance alone.

I have taken a few of the Coursera courses to (1) test out the format, and (2) satisfy some of my curiosity without wading through textbooks. Some of the courses are ill-taught and ill-managed, while others were sterling examples of instruction.

I would recommend checking out Dan Boneh’s course on cryptography, or Tim Roughgarden’s Design of Algorithms. Both give excellent lectures. And the homework reminded me of the level of problems I solved in upper level undergrad courses. Many were multiple choice problems, and so one could play process-of-elimination games with them. But the serious student can work out careful proofs before submitting. It is also possible to create problems that require numerical input.

In short, I expect it would be a lot of fun for a semester if you jump in with both feet.

I think that a component of any decision to teach or not teach such a course ought to be the role of MOOCs in the future of academic employment. In particular, is the MOOC to the teaching industry more like the record to the music industry? Or is it more like the MP3 to the music industry?

I’ll vote for “yes, you should” – I have a lot of respect for Clayton Christensen, so given that he thinks that higher ed is ripe for disruption, then if I were a professor then I would want to experiment a little bit with alternatives, to start getting a feel for the possibility space out there.

I studied math at Madison and so I would like to give my first thoughts about this. I was also a grader for the upper levels.

One mistake I made as student was to rely too much on the faculty to organize the material for me, and then to assign problems in order to complement the material. This isn’t a bad thing for the faculty to do, but in reality, most faculty there did no such thing. The notes-based courses often had too many gaps to make the topics completely coherent in the time frame, while the professor was unapproachable or had a bad attitude. Many text-based courses lacked insight beyond the text. Keeping up felt superficial, and it felt like I was creating a mathematical debt to myself. Feedback on problem sets was minimal, and usually not even by the professor, and communication and detail (like writing proofs in paragraph form) was rarely recognized, rewarded, or appreciated. In one higher level course a student was mocked after asking questions. I always sensed a general disregard for undergraduates and even graduates (“the job market will determine your grade”), and that perhaps part of the problem was that professors and grad students were stretched too thin. Why bother teaching business or engineering calculus when you don’t have much respect for these courses anyway? Let Business and Engineering do it. The math graduate courses I took were were also easy A’s, and so maybe the standards there are low or I’ve just had a bad experience or my perspective is partial. (I took about 15 upper level courses if it has any bearing.)

After that experience, I would prefer MOOCs for the senior-level courses with an emphasis on them being self-contained. So this is coming from the perspective that I know what it means to prove things, where feedback is perhaps less necessary all the time. I would still like some though, and also to be able to talk with an expert about the content occasionally, and so perhaps some options like scanning in handwritten work, multiple choice theory problems, or teXing up selected problems or a video conference could be made to work. I mean anyone can buy a book, read from it, and work through problems (it helps to have guidance in selecting them though), but feedback can be invaluable.

Eh, I don’t know.