Maybe online course aren’t going to remake the face of higher education after all.
After a fast start, reality seems to be closing in on the world of the massive, open, online courses that were supposed to replace traditional lectures and recitations and make free, or at least very cheap, higher education available to everyone. San Jose State University has slowed down a move to deliver introductory undergraduate courses through MOOC provider Udacity.
Udacity itself, one of several MOOC providers that have sprung up in the last couple of years, is refocusing its activities on corporate training. Sebastian Thrun, Stanford computer scienctist and founder of Udacity, told Fast Company, “We were on the front pages of newspapers and magazines, and at the same time, I was realizing, we don’t educate people as others wished, or as I wished. We have a lousy product. It was a painful moment.”
No surprise. I’m not surprised. I’ve been a skeptical enthusiast for online education since MIT started its Open Courseware initiative a few years back, and over the last year or so, I have enrolled in several offerings, mostly from Coursera, like Udacity, a for-profit provider of open courses. My experience has been a very mixed bag, but one that has taught me a lot about where the approach does and doesn’t work. A couple of general observations. First, the technology has a long way to go and on one seems to have figured out a completely effective way to deliver lectures on video. I’ve seen a number of approaches, from video recording regular blackboard lectures, to slide-based presentations in which the instructor only occasionally appears, to a course that used cartoony “virtual students” to asked questions in computer synthesized voices. None worked completely, though the last was the most annoying. Online lectures today remind me of the earliest days of television, when shows were “radio with pictures.” No one has quite cracked the medium yet.
Second, and more important, MOOCs seem to work best for those who need them least. Most surveys of those who enroll, in MOOCS and especially those who complete them–typically no more than 5% to 10% of the original enrollees–tend to be people who already have undergraduate degrees and often more. “If you’re looking to really move the needle on fundamental educational problems, inside and outside the United States, you’re going to need to help people reach the first milestone, which is getting their degrees to begin with,” Daphne Koller, a founder of Coursera, told The Chronicle of Higher Education. My experience is that MOOCs require very highly motivated students. It can be very easy to slough off, fall behind, and drop out. You have paid little or nothing, so there’s little at stake.[pullquote]If you’re looking to really move the needle on fundamental educational problems … you’re going to need to help people reach the first milestone, which is getting their degrees to begin with.”–Daphne Koller, Coursera[/pullquote]
Where MOOCs work. MOOCs work best for the sort of course that mainly consists of transferring information. I am finishing up a Coursera course in basic financial accounting, taught by Professor Brian J. Bushee of the University of Pennsylvania’s Wharton School. I signed up because I had just become treasurer of a non-profit and was embarked on a project of getting improved financial reporting for better decision support. Accounting isn’t the world’s most interesting subject and this wasn’t a terribly interesting class. But it gave me the knowledge I was looking for. Introductory accounting is a very good subject for online training: It consists mostly of a lot of sometimes arbitrary rules and how to apply them. And it lends itself well to effective multiple-choice evaluations. (This, however, was the course with the annoying virtual students.)
In a similar vein, I took an introduction to cryptography course from Stanford Professor Dan Boneh. Again, I though the course succeeded because it focused on straightforward information, not deep ideas. It was mostly aimed at teaching programmers what to do, and probably more important, not do, when implementing encryption, Some of the concepts were quite difficult and I thought that multiple choice evaluation was not terribly satisfactory; questions with longer free-response answers would have been much better. But the need to keep costs down drives machine-scorable evaluations.
The big surprise was a another Penn-Coursera course, Calculus: Single Variable, taught by Professor Robert Ghrist. This was not standard freshman calculus course; the course description suggested prior familiarity with introductory calculus. Ghrist took a fresh, almost idiosyncratic approach to a familiar subject (he focuses heavily on Taylor series and included a week on discrete calculus, a subject rarely touched in an introductory course.) Above all, he showed that multiple choice or numerical answer questions could be both interesting and challenging. This is the course to watch for anyone who wants to see how to do it.
A disappointment. My biggest disappointment was a course called Introduction to Mathematical Thinking taught by Professor Keith Devlin of Stanford. Devlin is a distinguished mathematician and an excellent lecturer, but this Coursera project was tedious. The lectures consisted mainly of him talking while writing, a sort of online version of the blackboard talk, but it did not work well for me. My biggest problem was that the problem sets, which I expected to deal with ideas, consisted mainly of slogging my way through endless truth tables. And Devlin’s attempt at supplement machine scoring with peer evaluations just did not work (In fairness, I took the course the first time it was offered; Devlin has taught it a couple times since and it may well have improved.)
There’s no question that the rising cost of higher education is a big challenge to U.S. society and that the inefficiency of the current system is a big contributor. But MOOCs as they exist today do not seem to be the answer, or at least not more than a small part of the answer.