This is an interesting and useful post. Bodong Chen first distinguishes between learning analytics and academic analytics (the former being directly concerned with teaching and learning) and educational data mining (the latter being more focused on the exploration of data from academic settings). He then outlines some areas of interest: first, the emphasis on big data in learning analytics, and second the need for it to consider the more nuanced aspects of learning itself. This leads to a discssion of the 'tensions': first, divisions based on different accounts of learning; second, the tension between learning and algorithms; third, agency and control; and fourth, the ethics of learning analytics. It's also worth viewing his Learning Analytics course.
Best takedown of the day: "The state of knowledge among people who have actually run large online communities is so far advanced beyond the research community that most research in this area is more amusing than helpful." The line comes from Michael Caulfield and he cites Clay Shirky on a pre-2011 truism about online community: "You have to find some way to protect your own users from scale... you can’t try to make the system large by taking individual conversations and blowing them up like a balloon." This of course is what we were also trying to do when we shifted our MOOC users away from the centralized Moodle platform and into their own blogs and communities. And as Caulfield says, "This is discovered repeatedly throughout history, and you could write a taxonomy of different techniques we use to protect users from scale." Image: Waltman and van Eck, community detection (worth a read in its own right).
I have a story I often tell. Suppose, I say, two people want to travel from Edmonton to Calgary. What's the best way to do this? Should they each get a separate car and race? Should they bid against each other for the one remaining car, with only the winner traveling? No, competition won't get these drivers to Calgary faster nor more efficiently. The rational thing to do is to share a ride. John Warner writes, "Competition works really well when the goal is to determine who is a winner and who is a loser and the winners benefit, receiving their tributes and rewards. When the rewards are outsized, or the punishment severe, truly terrible behaviors can result." He's right.
This is a good but not especially imaginative article on what to expect from mobile learning in education. For example, what factors will impact the use of mobile technologies? Bandwidth, instructor use and proficiency, and student proficiency, we are told. Well - yeah. Anything else? Or, for example, what can we expect in the future? We are told: location-based learning, augmented reality, wearable learning, internet of things, and 'apps' for learning. The same stuff, in other words, that we've been reading about for a decade. And yes, we are told to ramp up instructor training, secure leadership buy-in, and measure project results. Yawn.
Using a Moodle plugin, "instructors can turn their courses into a personalized game, where students complete course activities in the school's Moodle learning management system to gain skill points and advance their avatar through a series of objectives." Students take a pretest, receive an avatar, and then run through some 75 activities, working to earn 'skill points' which count toward getting their 'job' of choice.
Should games be hard-coded to run on specific hardware? Oculus is in a war against so-called hackers who are adapting their games to run on rival systems such as HTV Vive and Valve. Is this (as they claim) 'piracy'? Technically, yes, since the DRM must be broken. But using a legally purchased game on an alternative platform doesn't seem to be inheerently wrong. And even Oculus said it would be OK: "Last year, company founder Palmer Luckey posted on Reddit that he would not resist users' attempts to try out Oculus games on rival headsets." What does this whole episode tell us how Facebook - which owns Oculus - will operate in the future should it ever be able to lock users into its platform?