Educational Data Mining for Technology-Aided Formative Assessment

Liveblog notes from a CALRG seminar given by Ilya Goldin, from the HCI Institute at Carnegie Mellon University, at the Open University, on 12 Feb 2013, entitled “Educational Data Mining for Technology-Aided Formative Assessment”.

His PhD research was a system to help peer review. Mike Sharples met him at the Alpine Rendez-Vous.

(cc) paul bico on Flickr

Postdoc at Carnegie Mellon; PhD at Pittsburgh.

Continue reading “Educational Data Mining for Technology-Aided Formative Assessment”

MOOCs: Fail better

So another week, another high profile MOOC story.

A course on Coursera has had technical problems, which drew attention to shortcomings in the design of the course, which led to a lot of negative comment, and has culminated in the course being suspended. As many have noted, one of the big risks of working in the open is that everybody sees when you get it wrong. (Coverage in Inside Higher Ed, one participant’s account, and another overview of what happened with some interesting discussion in the comments. George Siemens has some interesting stuff to say, as ever.)

If that’s not bad enough, the subject of the course in question is “Fundamentals of Online Education“. Oh dear. Cue much mocking from the I-knew-MOOCs-were-rubbish camp, and wryly raised eyebrows from others.

Is this the end of MOOCs?

Of course not.

Firstly, there are plenty of other MOOCs, particularly on this subject. There’s the Open Learning Design Studio MOOC #oldsmooc, which runs until 13th March, or #ETMOOC, on Educational Technology and Media, which runs until 30 March. There will be more.

And secondly, failing isn’t a problem in and of itself. Everyone fails – or if they don’t, they’re probably not trying hard enough.* That’s particularly true of something like MOOCs where the whole point is to do things that haven’t been fully tried before. If you never get anything wrong, you’re being too conservative in what you’re trying to do.

No, the difficult bit is what you do when you fail. Coursera has taken the first step here: they’ve admitted there is a problem.

What they do next is what really matters, and will be the real indicator of their quality and potential. If they learn, and get better, the chances are they’ll keep doing that until they’re really very good.

As Samuel Beckett put it “Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.”

I do feel sorry for Fatima Worth, from Georgia Tech, about whom some rather unpleasant things have been said. The ghastly maxim “success has many fathers, but failure is an orphan” seems appropriate. Sure, she made some mistakes which with hindsight look obvious and glaring – but I don’t think they were quite *that* obvious and glaring beforehand. What applies to Coursera applies to her as an individual too: failing is not a disaster. What happens next determines whether or not it turns out to be a disaster. I hope she moves on and in ten years can open keynotes by using this as a self-deprecating anecdote of a hiccup from the early pioneering MOOC days.

At the time of writing (Tue 5 Feb 9.30am GMT), the course is listed with the next session ‘Date to be announced’.

I still think MOOCs Are A Good Thing.**

* Or they work in a safety-critical field, such as the nuclear industry or aviation, where learning-by-failing would entail unacceptable losses. Those fields have their own disciplines and techniques for wringing the maximum learning out of what failings or near-failings do occur, and these are indeed excellent. But they are also very expensive – too expensive for fields where nobody dies if you mess up.

** I’ve decided that to be more effective in communication, I should reduce my thoughts on major issues in my field to pithy oversimplistic slogans, which I will repeat frequently. I’ve already got “It’s Not A Power Law (Probably)”, and I hereby nominate “MOOCs Are A Good Thing” as another. I might also add “Here Comes Private HE”, although that’s a bit UK-centric, and it’s coming a little slower than I expected.

This work by Doug Clow is copyright but licenced under a Creative Commons BY Licence.
No further permission needed to reuse or remix (with attribution), but it’s nice to be notified if you do use it.