Back in early November Dr. Mike Johnson, Director of Institutional Research at Dickinson College, gave a stellar web presentation on how he used RI Analytics to tackle a notoriously challenging issue in higher ed: Retention Modeling. Who is leaving our college? Why? How do we pinpoint those likely to leave and possibly keep them around? How do we adjust our recruitment and more general institutional practices to address what we’ve learned?
Dr. Johnson is something of a legend in the Rapid Insight world, having once used Analytics to predict the size of his small institution’s incoming class within six candidates. His presentation does not disappoint. He explains the reasoning and methodology behind the Dickinson retention scenario, and how Analytics was uniquely helpful in cracking yet another tough research problem. Watch the entire webinar here.
The presentation is exceptionally thorough, detailing the subtle philosophy involved in presenting the issue, the available datapoints and how they may or may not be useful, and the impressive side effects of a problem well solved (such as predicting first year GPA in addition toretention/attrition likelihood). Also, Dr. Johnson does a great job of walking through the Analytics software and explaining the modeling process.
If you are you involved in predictive modeling, quantitative research in higher ed, or are just simply curious about how this stuff works, the video is definitely worth a watch.
-Ryan Moran 12/14/2010
No comments:
Post a Comment