This is a very interesting interview with Matthew Leavy, the Chief Executive of eCollege. Colleges and universities collect an incredible amount of data, whether in classroom evaluations, student and graduate surveys, professional development forms and even tenure and promotion applications. Unfortunately, too often this data sits on the shelf in its own little silos. We need to do a better job of freeing up that data and exploring ways to use the data to improve education. Data mining - although not the only way to do this - would be a great start. Data Mining as a Teaching Tool in Distance Education - Chronicle.com
Q: How have analytics and data mining shaped the services offered by eCollege?
A: That's a very critical piece of our strategy, and we think it is important for supporting online students and faculty. The holy grail of online education is to be able to provide personalized instruction for every student. Every action is captured in the data. We can learn how students learn and what makes them successful. So you might learn, for example, that if one student spends more than 10 hours in the discussion thread, they're more likely to fail out of the class. That's data you can use. "Triggers for intervention" is the jargon in the industry, and eCollege provides the tools to identify these triggers.
Q: Have analytics and data mining caused eCollege to develop new products?
A: We keep evolving the tools that we currently have. We just built something around Cognos, an IBM company that focuses on business intelligence.
Q: Have professors been using analytics and data mining to shape their curricula?
A: I certainly hope so. The great promise of working with a lot of data is that you can then evaluate if students using one curriculum or in one teaching environment do better than students do in another curriculum or teaching environment, and you can use that to constantly improve your approach.
No comments:
Post a Comment