Thursday, July 28, 2011

The FUSION Buzz - All About Analytics

While "attending" the D2L Users Conference, FUSION, in Denver this month, there was no other topic mentioned as much as the Action Analytics project that is currently under development. During the Product Launch presentation, John Baker and Jeremy Auger brought Brian C up on the stage to run through a little demo.

I was expecting to see the continued build out of a project that was started at MnSCU a couple of years ago, and that's exactly what I saw. I had the pleasure of working on this project with Al Essa and a few other folks while both Al and I were employed within MnSCU. Several things have changed since then: Al now works for D2L, I now work for nobody, and MnSCU killed the development project of what was easily the most innovative thing that has been done there in many years.

What we called Action Analytics at MnSCU, is called the Student Success System at D2L. However, before I go on, let me make it clear that although there are many similarities between the two projects, the D2L project did not continue the work started at MnSCU; they created a similar, but different, analytics prototype.

The demo was shown on a touch-screen (probably an iPad) tablet. The visuals begin below with what you might think of as a "seating chart" for an online classroom. For each student you get a visual reference to the progress that they are making in the class - currently shown as the small icon in the lower right corner of each student photo. Of course, the visuals will likely continue to evolve as the projects proceeds.

To learn more about a particular student, you will simply click (or touch) the photo to drill down to more information. Or, as the photo below illustrates, you can filter the class enrollees to see which of them are at-risk, or doing great, or whatever.

In this case, Bobby D isn't doing very well in the class. The next view will provide information about the how well prepared the student is for college-level work, how well prepared he is for this particular class, and similar things such as how much the student has engaged with various course tools or other requirements.

The tool will also be used (if desired) to help predict the likelihood of student success based on comparisons with other students in the class or with previous students in similar sections. Much of this predictive use of the data is still being researched (generally speaking, by everyone doing this type of analytics work) and the final functionality or utility of these predictions is not certain.

Once a student has been identified as being at-risk of non-completion, failure, or other negative consequences; various possible remedies/interventions/reactions are possible that can be initiated by the instructor or advisor, or whoever else is using such a tool.

There are still many things to be worked out before this SSS is ready for prime time, but the D2L team appears to be making tremendous progress down this road.

Now would be a very good time for colleges and universities to start thinking about what policies and procedures they need to put in place in order to effectively use this type of tool. Who will have access to it, what actions can they initiate with or without approval of someone else, should students be allowed to opt-out of such a system - and about a hundred other questions that will come up. I know someone who could help with that project.

I am proud of the work that we did at MnSCU on a similar project, and I am excited to see the prime-time version of this tool when D2L is ready to release it. Different timelines for release have been bandied about, anywhere from 3 -12 months (Oct. 2011 to July 2012).

I've looked at several other data analytics projects that use data from an LMS. In my opinion, this is by far the best I've seen. Yep, by far.

Monday, July 18, 2011

Blackboard Client Exodus

Last week I sort of attended the D2L Users Conference (D2L11), also called FUSION, in Denver, Colorado. I love Denver and the surrounding area, but since I grew up in Cheyenne I've never been too impressed by the whole "Mile High City" stuff. Cheyenne is over 6000 feet in elevation, which I believe is higher than one measly mile.

I was a presenter at the Unconference (emceed by Kyle Mackie) that preceded the main conference, and I closed the evening with a little ditty titled "If I had a Beer..." This unconventional presentation was half-serious and half not-so-much. The basic premise was that a nickle is not really worth very much, so the whole "If I had a nickle for every time I've heard that" thing just doesn't seem to add up for me. However, if I had a beer for each of those times - then I'd really have something of value.

For example - the bit started with the following.

  • If I had a beer...
  • for every stinking time I've
  • clicked a link on the INTERNET
  • only to find that it was a link to a WORD document,
  • then I'd have a whole boatload of beers.
That was followed by a short rant about how stupid it is to put non-web documents on the WORLD WIDE WEB when it's just as easy and much better to put up a web-friendly document instead. There were about 10 of these little snippets of things that either make me crazy or make me go "huh?"

The final little snippet of the evening was a little shot at Blackboard, the recently acquired former-juggernaut in higher education. It went a little something like this, although there were graphics included: (BTW, these were not intended to rhyme, nor did they do so)
  • If I had a beer...
  • for each of the Blackboard clients,
  • that is currently looking elsewhere,
  • then I'd never go thirsty
  • with at least 300 beers in my fridge.
Later that evening, someone highly placed in the LMS industry told me that my number seemed a bit too low. Their research has indicated as many as 700 Bb clients are currently considering a switch to a different LMS.

Seems to me that 300 is a really big number when it comes to potentially lost clients. Clearly 700 is an incredible number. It's also true that Blackbeard won't lose all of the captives (err, clients) - but it begs the question of what percentage of these clients are they likely to lose? If the past few years are any indication, my guess is that they will lose over 50% of the customers who evaluate BlackAngel's offerings in comparison to the competition.

The pic shown below is a view of part of the enthralled audience at the Unconference.