A few weeks ago, I attended a symposium on learning analytics, jointly sponsored by NERCOMP and Educause ELI. It was a day well spent. The morning session consisted of presentations on Learning Analytics; the afternoon session was hands-on. Each team was asked to design an application that would do some form of learning analytics. What was at stake was a $10,000 check from a venture capitalist (all in good fun, of course). Each team presented their application, at the end, and the group voted. My team developed an application that would follow students eyes as they read course materials, as they interacted with the course. The thought was it would reveal the effectiveness and use of various course materials. (It tied for first place).
What I came away with was lots of questions, lots of ideas regarding the use of learning analytics. The power of learning analytics is that they allow us to truly understand our students. They help us target intervention with greater accuracy. One institution shared that information gathered from their LMS showed a very distinct correlation between failure rates, and dates of first logins to a course. This knowledge allowed them to develop interventions that could be put into effect the first weeks of the course.
The presentation that most interested me, was on the development of software that could “read” essays and determine the extent of comprehension. I can see this as an extremely useful tool, most especially for MOOCs and other courses with large enrollments.