Video Lectures are Good: MOOCs for science dissemination and an open access to education
– John Lees-Miller reporting from Trieste
Hello from the Abdus Salam International Centre for Theoretical Physics in Trieste, Italy! 1 The ICTP supports science and science education in the developing world, and we got an invite to one of their workshops.
As usual, I learned much more from the workshop than the workshop did from me. I was invited to talk about writeLaTeX and how it makes collaboration easier for scientists and students around the world. But this post is about all the other great stuff at the workshop. The main focus was on e-Learning and Massive Online Open Courses (MOOCs). I've done four MOOCs now, 2 so I have some first hand experience, and it was great to hear what the experts had to say. So, here are some of my key takeaways and thoughts.
Massive, with Video
There are two new things about ‘Massive Open Online Courses’. The first is ‘Massive’. In 2011, Thrun and Norvig's Introduction to AI course attracted 160,000 students. That's a really big number. They taught more students in that one term MOOC than they could have in many life times of conventional teaching. And they did it for very little money. MOOCs use the web to educate at scale.
The second new thing is that video now works on the web. It was not so long ago that web video meant pixelated and matchbox-sized video with tin-can sound. Now well produced, high definition video lectures with clear audio are the backbone of every MOOC. It also helps that we now have mobile devices that can play these videos — I watched most of my MOOC lectures on my phone while on buses and trains. Granted, I did have to watch some of them again later when I could pay closer attention and take notes.
Video Lectures are Good
That's one of the great things about video lectures: you can pause them and replay them. And you can even search them so you can find the relevant bits again when revising. Another simple advantage is that the length of a video lecture is be determined entirely by its content; most MOOC lectures are very short and focused — often less than ten minutes, after which there is some form of quiz or review.
Of course, what you can't get from video lectures is interaction with your teacher and peers. The best lectures — and lecturers — I had were also the most interactive. In MOOCs, interaction still happens, but it happens separately from the lectures. It comes mainly from message boards and study groups. I met some great people at a study group for Daphne Koller's graphical models course.
Who Takes MOOCs?
It seems that a lot of people like me take MOOCs, and unfortunately we skew the statistics a bit. Michael Schatz at Georgia Tech presented some figures from the introduction to physics MOOC that he's been running on Coursera, and one of the most interesting ones was that over half of his enrolled students already had science degrees. 3 So, they presumably weren't there to learn basic physics; instead, they were probably there because the course sounds amazing: students use smart phone cameras and video analysis software to study the physics of motion in the real world. Sign me up!
The presence of all these highly educated tyre-kickers makes it even harder than usual to interpret the numbers. In particular, MOOCs have very low apparent completion rates; typically less than 10% of the students who sign up actually complete the course. But, it's hard to say at present whether this is a ‘drop out’ problem or a ‘drop in’ problem — if so many people start without seriously intending to finish, is a low completion rate really a bad thing? It's a bit early to say.
Big and Small
Another interesting question about where MOOCs are going is, will there be a MOOC monopolist? There is certainly a tendency toward monopoly on the web, and in technology in general — Facebook, Google, Microsoft. It would be dangerous if a single company controlled how everyone learned, say, physics, and even worse if it controlled how everyone learned, say, history.
Fortunately, this does not appear to be happening. There are certainly economies of scale, but network effects do not appear to be particularly strong. There are several strong players in the MOOC space, and more are still entering. Moreover, one of the most interesting and encouraging parts of the workshop was to see how teachers and institutions around the world, and particularly in the developing world, are using and adapting the techniques and technologies behind MOOCs to the needs of their own students. Here's hoping their work leads to a future where everyone learns cheaper and better.
1 This post was (mostly) written three weeks ago. I'm not in Trieste any more, sadly.
Thanks to Marco Zennaro and Enrique Canessa for organising the workshop and inviting us, and to everyone who came, and especially @DonaldClark, @sebschmoller and Michael Schatz, who had the most influence on my views and this post. Thanks also to Makoto Innoue and the crew at Bethnal Green Ventures for reading drafts.