Higher ed data guru Nate Johnson of Postsecondary Analytics, LLC, has produced a new online tool that allows users to mine up-to-date (academic year 2010-2011) degree completion data by level of award, state, and type of institution. Check out the article about it in yesterday’s edition of Inside Higher Ed here and check out Dr. Johnson’s degree completion tool here.
I was able to produce the following chart for my home state of Louisiana:
Depressing, yet informative…..but that’s the subject of a whole other post!
While the data is nothing novel (it’s simply preliminary data submitted to IPEDS), it is much more timely than the published data on which most policy makers make decisions. As I scan my office at this very moment, there are piles of reports on my desk and binders of reports on my shelf containing the analysis of data that is 3 to 5 years old. I routinely take these reports with me to meetings and decisions are based upon what has happened in the past, in summary statistic form.
And while Dr. Johnson’s tools bring us to the present (a much needed improvement indeed), I ask, why are we, in higher education, constantly looking in the rear view mirror?
Harnessing the power of predictive analytics, private industry studies patterns found in historical and transactional data to identify risks and opportunities, and to make decisions about the future. Modeling allows analysts to capture the relationship between variables, isolate predictor variables and more or less predict the future. This has worked to great effect in the marketing and insurance fields.
Predictive modeling is most certainly taking place in higher education. I’m not here to argue that it’s not. But it’s taking place at the campus level in admissions and development offices, is siloed in education think tanks, or is financed by philanthropic foundations with dubious policy aims. It has not, however, taken hold at the state or federal levels. Or at least it has not taken hold in any holistic fashion. Legislators, governors and higher education governerning bodies are still making policy decisions for the future based on dated summary statistics, with little analysis of the predictive capability inherent in those statistics.
Paul Fain’s article, entitled “Big Data’s Arrival,” in the February 1, 2012 edition of Inside Higher Ed highlighted the findings of a predictive analytics project funded by the Bill and Melinda Gates Foundation (to the tune of a $1 million grant). Fain commented that these kind of predictive analytics have “long been embraced by corporations, but not so much by the academy.” It should be noted that the project’s findings flew in the face of long-held assumptions about at-risk students. If we were to truly begin to look at things through the prism of prediction, how many more long-held assumptions would be put to the test?
Shouldn’t the onus be upon elected officials and higher education policy makers, not privately run foundations, to study such matters and to study them in a way that allows us as a collective whole to make decisions about the future based upon reasoned predictions?Instead, we continue to count the beans in aggregrate form, calculate year over year percent changes, and report in summary fashion. We pay little attention to the underlying relationships and how those relationships can inform the path ahead.