Fortnightly Mailing

Udacity's approach to course improvement

Small "on re-reading this" edits almost immediately after publication.

You may have read this fierce critique of one of Udacity's courses by Daniel Collins, an experienced  and articulate community college maths teacher who will for sure appreciate at first hand the challenges faced by students who do not take naturally to learning mathematics.

What particularly caught my eye was the unusually open way in which Sebastian Thrun, founder of Udacity and teacher of the Statistics 101 course has responded to the points made by Collins, and the gracious and constructive way in which Collins acknowledges this.

All three posts are well worth reading, along with their developing discussion-threads.

This is the kind of relatively open and self-critical dialogue that needs to be taking place between people who understand the problems and possiblities of the different kinds of mass courses that are currently under development, and who understand their fields from the point of view of how people learn. And if I am judging the underlying quality of a mass course provider, then its openness to criticisms and the extent to which it is committed to incremental improvement in its provision is a key consideration.

The chances of edX, Coursera, Udacity et al reaching "Facebook numbers" should not be lightly discounted

I mentioned EPIC 2020, a US-focused Call to Action to compel universities to 'decouple the delivery of content and assessment through a “test out” option' in Snippets from 14 July to 3 August.

Today, via science librarian John Dupuis, I found this interesting critique of / commentary on EPIC 2020, by Justin Marquis.

Though I do not agree with Marquis's assertion that MOOCs currently have a serious and probably inherent flaw in their inability to help people develop their creative and innovative capacities, the piece is worth reading in full.

It is also worth noting the comment on the piece by Bill Sams, originator of EPIC 2020.

"Thank you for the excellent review and discussion of EPIC 2020. My objective in producing EPIC was to create a piece that would cause people to consider and discuss that there are dramatic alternatives to the traditional education system. Given the 25,000 views from 83 countries I am satisfied that I have made a small contribution to what hopefully will become a lively discussion. On a side note as to the timeline: In about seven months Coursera has enrolled one million students. Facebook took ten months to achieve the same level of members. Five years later Facebook is in the neighborhood of one billion members and has a capitalized value of $41 billion. The chances of Coursera and edX reaching similar numbers should not be lightly discounted."

Mark Guzdial's talk with Sebastian Thrun

Mark Guzdial has written this informative and sensibly low-key piece about Thrun's thinking, which benefits greatly from Guzdial having done a lot of prior thinking about teaching and learning computer science.

Online learning - helping learners create learning

Links to Koller and Horowitz talks added 21 August.

At the foot of this post is an 18 minute talk by Peter Norvig at the Google 2012 Faculty Summit on 26 July. In it Norvig reflects on what he learned from developing and running last year's "Stanford" online AI course (in which I participated), making links as he goes with the widely applicable "Theory and Research-based Principles of Learning", from Carnegie Mellon University's Eberly Center for Teaching Excellence.

Two othe things stood out for me from the talk.

The first is the reference Norvig makes to the challenges posed for course design and operation by the "dynamic range" of an online openly enrolled course, by which he means the wider range of capabilities and experience than would be found on a course where there are strict admission requirements.  Norvig is not claiming that this idea is new, but it is good that it is getting attention.

The second is this very striking quote from the polymath Herb Simon:

Learning results from what the student does and thinks and only from what the student does and thinks. The teacher can advance learning only by influencing what the student does to learn.

which squares strongly with something that Dylan Wiliam said at the 2007 ALT conference (where Peter also spoke):

Learning power is a concept that Guy Claxton has put forward. The key concept here—the big trap—is that teachers do not create learning. That’s true—teachers do not create learning, and yet most teachers behave as if they do. Learners create learning. Teachers create the conditions under which learning can take place.

(The full transcript of Dylan's talk is available for download [PDF]. Other talks from the Google 2012 Faculty Summit are also available: Daphne Koller; Bradley Horowitz - hat tip to R Seiter. Horowitz's talk about Google +, for which he is responsible at Google, is particularly interesting.)

Daphne Koller's London talk “The Online Revolution: Education for Everyone”

I made some small amendments to this post on 14 August.

Here is Daphne Koller's 2 July 2012 University of London talk “The Online Revolution: Education for Everyone”, the focus of which, according to the organisers, was particularly on "the pedagogy/platform that sits behind Coursera".

See also:

1. the comment below from Laura Czerniewicz from Cape Town University on what Daphne Koller says in her talk about South Africa;
2. this link to MOOC pedagogy: the challenges of developing for Coursera.

As an aside, it is a pity that the recording's dovetailing of Daphne speaking with her powerpoint presentation and its video-clips is a bit patchy. As a partial remedy for this, you may find it helpful to have on screen the slides used by Daphne's collaborator Andrew Ng during a similar presentation he made in Toronto around the same time, though these do not contain the video clips.

MOOC pedagogy: the challenges of developing for Coursera

ALT has just published a timely and informative "from-the-sharp-end" article by Jeremy Knox, Sian Bayne, Hamish MacLeod, Jen Ross and Christine Sinclair from Edinburgh University's MSC in E-learning Team about their experiences of and reflections on developing the E-learning and Digital Cultures Coursera course. Related to this from the ALT Open Access Repository, you can access a presentation PDF by Andrew Ng and Daphne Koller of Stanford University and Coursera. "The Online Revolution: Education at Scale" was used by Andrew Ng for his 23 July 2012 Invited Talk "Teaching Machine Learning to 100,000 Students" at the 2012 conference of the American Association for the Advancement of Artificial Intelligence. Note that a June 2012 TED talk by Daphne Koller and a July 2012 talk at the University of London cover similar ground.

[Disclosure: I have a minor paid relationship with ALT.]

New insights into Udacity, its learning and business models, and Thrun's thinking

This 3 July 40 minute interview by Jason Calcanis with Udacity founder Sebastian Thrun is, how shall I say? very "West Coast", even setting aside the advertisements.

I found myself feeling greater empathy with the interviewee than the interviewer, though you do have to admire Calcanis's interview technique, apart from his tendency to interrupt a bit too much.

Thrun has a very clear vision; his heart is in the right place; and he has understood in a deep and serious way how online learning, done on a really big scale, and organised and supported in the right way, can and will change things for the better. The kinds of changes that Thrun envisages (to models for learning, to what it means to be a teacher, to the educational establishment) will -- if they are done right -- be worth it from the point of view of learners and the world at large. But if you were now setting out on a career as a teacher, especially in post-compulsory education, wouldn't you have to take seriously Thrun's points about the fall in the proportion of the workforce in the developed world involved in farming, once modern mechanised methods took hold?

[By way of a counter-argument, see Jordan Weisman's Why the Internet Isn't Going to End College As We Know It in The Atlantic, via Mark Guzdial.]

Edinburgh University bites on the "MOOC" bullet with Coursera

Don't be put off by the slightly stodgy tone in parts of this just-released promotional video from the University of Edinburgh about www.ed.ac.uk/moocs. Instead, listen carefully to what Stanford's Daphne Koller has to say about scale and formative assessment in Coursera's new "breed" of free on-line courses, as well as to the comments from Vice-Principal Jeff Haywood about Edinburgh University's approach to quality assurance. [See also coverage by BBC, Guardian, The Times Higher, Inside Higher Ed, Downes. ]

Of possible interest to readers of Fortnightly Mailing will be one of Edinburgh's Coursera courses E-learning and Digital Cultures, taught by Jeremy Knox, Sian Bayne, Hamish Macleod, Jen Ross, and Christine Sinclair. E-learning and digital cultures will "explore how digital cultures and learning cultures connect, and what this means for e-learning theory and practice". [On 8 August, ALT published MOOC pedagogy: the challenges of developing for Coursera, Jeremy Knox, Sian Bayne, Hamish MacLeod, Jen Ross and Christine Sinclair.]

The emergence of mass on line tutoring systems (MOTS)

Epigeum is a spin-out company from Imperial College London. Co-founder David Lefevre sent me this piece about mass online tutoring systems. He wrote it partly prompted by hearing a talk - "one of the most impressive presentations I have seen in our field" - by Daphne Koller (co-founder of Coursera), on 2 July at the University of London Computer Centre.

Lefevre is not making a new point, but he gets to the heart of the matter, in this snippet:

Human tutors are present but the delivery system allows them to operate largely at a meta level and therefore teach many more students than is possible via more traditional methods. .....  As anyone who has taken a course on these systems will attest, the learning experience does not feel impersonal. The effect is a rich, engaging experience far removed from the solitary browsing experience provided by OpenCourseWare.

(I wonder if the MOTS acronym will stick?)

Peter Norvig's TED talk reflecting on creating and running the online AI course

[Added 9 July 2012: The discussion on Hacker News prompted by this post by Colin Wright is well worth reading.]

Peter Norvig is Google's Director of Research. Here, finally, is a recording of Peter Norvig's February 2012  six minute TED talk about his and Sebastian Thrun's 2011 mass free online computer science course about artificial intelligence (in which I took part as a student and from which I wrote weekly reports throughout - to find them scroll down here). It is interesting in its own right. It also shows how much can be conveyed in only six minutes (though post-talk questions and discussion would have added a great deal).

Note Norvig's conclusion:

"....the most exciting part of it is the data that we're gathering. We're gathering thousands of interactions per student per class, billions of interactions altogether, and now we can start analyzing that, and when we learn from that, do experimentations, that's when the real revolution will come."

[Below the video I have pasted the transcript of the talk.]