Even critics of massive open online courses, better known as MOOCs, shouldn’t deny the value of the student data those courses generate. Teachers can only gather insights into how engaged students really are with the material and how well they’re understanding it if they’re using a platform designed specifically to capture that data. MOOCs do this very well, and now University of Michigan meteorology professor Perry Samson (who also co-founded Weather Underground) has developed software to let his peers in lecture halls do the same.
The platform is called LectureTools, and it has some obvious benefits around helping ensure students engage with a course more than is naturally possible in a room full of 250 people. While class is in session, LectureTools lets professors quiz students using a variety of different formats, lets students submit questions and note when a slide confuses them, and even lets professors teach the course remotely if need be. At any time of day, during an afternoon lecture or at 3:00 in the morning, students can access the slides and other course materials and type notes, draw diagrams and generally engage with the material as they wish.
Everything is stored in the cloud, so professors and students can access their work whenever and wherever they need it. “I have got to know in the first two weeks are you falling through the cracks and how am I going to deal with that,” Samson told the audience during a session at the Strata conference on Wednesday. Tracking students’ progress daily rather than waiting until a mid-term exam helps him do that.
However, Samson explained, LectureTools also has promise beyond the scope of any given student because it lets him collect and analyze data he never would have been able to in the past. All the notes the students type are saved and the text analyzed (generally by creating a word cloud right now, so others can see the concepts they should have caught from the lecture) but also much deeper sociological data. For example, Samson begins every class by having students quantify how they’re feeling physically and emotionally, so he can correlate well-being with performance in class.
Going forward, all the data LectureTools is collecting should let the team working on it automate the feedback process so students can figure out if they’re on the right track even while they’re studying late at night. Samson also said he’d like to open access to the data to students and professors via API, so creative types figure out their own ways to learn from it in new ways. He also thinks releasing the the data to the research community, anonymously and in aggregate, could uncover a lot about how students learn and study.
We know that most MOOCs are capturing this data and that LectureTools is, too. I have to assume that most, if not all, other smart attempts at online education — such as the University of California, Berkeley’s online data science master’s degree program — are, as well. Because there’s a big difference between just freeing up content like video lectures and actually building platforms designed to improve outcomes. I’ll have a chance to ask about the UC Berkeley program at our Structure Data conference next month, where I’ll be interviewing AnnaLee Saxenian, dean of the university’s School of Information.
Even if analyzing data isn’t an elixir for what ails the university experience, any improvements it can effect are meaningful. Because right now, Samson said, especially when it comes to helping students learn in the large lecture courses that generate so much money for the schools, “Universities are failing, and failing miserably.”