Data mining online courses

What if we had the tools to identify that most of the students are missing the same exam question and aggregate that data across many sections? What if we could tell that, while only 10 students in a course section are missing the same question, a far larger number are missing it across all the sections?

We actually have the ability to obtain that type of data now through online education. If we design our online courses (or components of blended courses) well and are consistent in what and how we measure, the law of large numbers and the analytics of our courses can help us identify areas where students need help.  Then, we can allow students to access that help in the way that makes sense to them.

Don’t take my word for it.  Listen to Daphne Koller, who has enough research experience and academic credentials to be more than credible when she describes the powerful analysis tool that we have in well-developed online education (described about 15 minutes into her video).  Technology is not only allowing us to reach more people, it’s allowing us to understand how well we are reaching them.

One of my favorite examples of engaging students differently online than we can in face-to-face lectures is Koller’s description of what happens when she asks a question in her class.  She reaches the point in her lecture where she wants to ask a question, but most of the class is still writing the last thing she said.  One student at the front answers the question before most have realized that question has been asked.  What if they could all have the time to recognize the question, formulate their answers, and then respond?

And what if we could then analyze how many got it right?  And allow the ones who got it wrong to choose their methods of review and/or redo?

I’d like to hear your thoughts.

Leave a Reply