As Salman Khan admits in his book One World Schoolhouse, he was actually lucky that YouTube initially had a 10-minute limit on the length of the videos he posted. The shorter, the better!

YouTube has great stats on audience retention, by the way.

Check out the results for one of your own videos:
http://support.google.com/youtube/bin/answer.py?hl=en&answer=1715160

Look at how long people spend on Wikipedia, on average:
http://www.alexa.com/siteinfo/wikipedia.org
then click Time on Site:
Yesterday 5:06

Clearly, MOOCs provide more audience tracking than was ever possible in traditional media. The value of user data might provide commercial stability as they hone in on what works best for students, faculty and university partners.

Here is a summary of the MOOC completion rates research:
http://www.katyjordan.com/MOOCproject.html

and the report on the MOOC with the highest rate:
Functional Programming Principles in Scala: Impressions and Statistics
http://docs.scala-lang.org/news/functional-programming-principles-in-scala-impressions-and-statistics.html

“We had more than 50,000 registered students— an unfathomably large number in the context of traditional teaching. While large, that number doesn’t tell the whole story; as is typical for a MOOC, a statistical majority of those students participate no further beyond watching a couple of videos to find out what the course is about. Of the 50,000, about 21,000 students participated in the interactive in-video quizzes that are part of the lectures, and a remarkable 18,000 unique students attempted at least one programming assignment. A whopping 9,593 students successfully completed the course and earned a certificate of completion— that’s an incredible 20% of students, which blows the average 10% rate of completion for MOOCs out of the water.”

Anyway, to me, the word “university” has changed. It used to mean a place you could go in which the entire universe of knowledge and potential was available to you. Now we have direct access to that universe and yet we still need we connecting points, guidance and shared experience. How can knowledge, practical know-how, and collective intelligence be applied for each student, at their point of need?

(And I’m hardly anti-MOOC: I’m a computer science professor, in charge of online education at Berkeley, instructor of Berkeley’s first MOOC ever which has been offered on Coursera and EdX, and I’m on EdX’s technical advisory board and helping to run conferences and workshops on the role of MOOCs and online education generally.)

First, there is no evidence that MOOCs work particularly well *for college students.* Most enrollees in our MOOCs, and those of several of our colleagues, are working professionals with several years of industry experience under their belts. They are arguably better at personal time management than first-year undergrads. They approach the course, in many cases, with more background; for example, our Software-as-a-Service course uses Rails, and although that framework is new to most people, the practice of programming is not.

Second, I’m a bit uneasy with the current Udacity pilot at SJSU because the college has apparently delegated the entire course delivery experience to Udacity – assessment, TA support (phone/web), and so on – with local instructors participating only in the design of the course, not its administration. That’s a small step away from simply hiring those instructors as one-time contractors for course design, cutting out SJSU altogether. (And before you object that that might be the right thing to do, bear in mind all I’m saying is no one knows yet if that’s a good idea.)

The enrollment issues at community colleges (and to a lesser extent Cal States) are real, but they are not uniformly distributed. Those who appear to be most affected are low-income students, often students of color. Distressingly, these are the students most likely to lack the computer and broadband facilities needed to avail themselves of the potential benefits of MOOCs. (The enrollment issues for gateway courses are barely an issue at the UC’s, whose time-to-graduation is a shade over 4 years systemwide.)

There is a great deal of pedagogy research to be done on MOOCs – at Berkeley our approach is to blend the MOOC with in-classroom experience, so the rest of the world gets the MOOC but the Berkeley students get a combination of MOOC and live instruction. We’re hoping to use this approach as the basis for doing serious research on what MOOCs can and cannot accomplish, whether standalone or in conjunction with “on-the-ground” instructors.

But until some of that research is done, I’m nervous that legislation is racing ahead of pedagogy, especially in an area where there is undoubtedly lots of money to be made – and that’s fine, but I just don’t want it to be made at the expense of the student experience.

Great post. Keep them coming.