During this academic year I will be working on a project called Action Analytics with the Office of the Chancellor at the Minnesota State Colleges and Universities (MnSCU). I will still retain my position at Lake Superior College, but part of my time will be devoted to this project.
Dr. Linda Baer of MnSCU was a recent guest blogger on Donald Norris’ blog Linking Analytics to Lifting Out of Recession. Dr. Baer described several data analytics initiatives that are ongoing within MnSCU as serving diverse populations and also the underserved populations. As a more recent step, she describes the following: “Linking Activities and Metrics. The second foundation step is to begin to link these many activities. The technology and infrastructure capacity of the system was supporting a large data warehouse with selected data mining capabilities. The dashboard enabled full display of campus and system accomplishments. Yet, the question remained: What activities were contributing to the most student learning and student success?”
Boosting Analytics and Predictive Modeling. Dr Baer continues with the following: “Realizing that the next big step was to develop analytic and predictive modeling capabilities … using national, state and local data, we could develop student information dashboards so each faculty member could review where the student was academically and then advise the best academic choices for ongoing success. Curricular assessment could be made to see what components of course learning worked for students and what needed more emphasis so tutoring and advising could further align best learning experiences to accomplish successful learning.”
The desired end result is to create an informational system that can be used to provide students with timely advice about how best to succeed in their current courses and in their more long-term programs oif study. To that end, I am working as part of a team to create a proof of concept for this project. We are analyzing sudent engagement data harvested from their activities inside Desire2Learn during the spring 2009 semester. Early results have favorable indications that the student engagement data might indeed be useful as a predictor of success in the course, although that analysis is not complete at this time.
Stage one is to create an advising model where advisors would be able to quickly see which of their advisees might be at risk academically and to help make suggestions as to which type of actions (interventions?) might be useful in bringing the student closer to a successful outcome. This work is ongoing and I expect to have much more to report about it over the next weeks and months.