It’s all over Twitter today:
IBM and Desire2Learn Take On Education Data Challenge bit.ly/KMa0Er #edtech
—
T.H.E. Journal (@THE_Journal) May 07, 2012This is an interesting step in data management and analysis and predictive analytics in education. I think what this brings to the front is that analytics is very much a business application, and that applying it to learning changes the education game, considerably. I’m not saying that it’s all bad. In fact, I think there can be a lot of good in it. With any discourse around this, I feel compelled to throw in a couple “let’s be cautious” and “consider the implications” type comments. I’m excited by predictive analytics. They’re neat, and I love all the pretty graphs. Should IBM and D2L be making conclusions and interventions based on the mass of data available? What are the implications and potential pitfalls of having edu-business-borgs making conclusions and giving advice? Consider what can’t and shouldn’t be counted. Let’s be sure to question the robots when they make decisions based on logic like this:
The best analytics applications, IMO, either provide the information to a knowledgeable human like an advisor or the faculty teaching the course, to make a decision whether to act or not (like D2L’s from my understanding of Alfred Essa’s presentations). OR they provide simple information directly to students/users in a way that’s easy to understand, for the user to decide whether to act on it (like Signals from Purdue).
Really, the only way it should be like the Monty Python witch clip would be for low stakes stuff, like Netflix recommendations or Amazon suggestions. We’ll see if that actually bears out in the implementation though, since educational admins are very fond of automating “everything”.