Graduation rates, student loans and big data

I always feel a strong connection to the cover stories of CIO’s monthly digital magazines, but September’s topic really hit home. Improving graduation rates is an issue all colleges and universities grapple with. From a business perspective, mediocre graduation rates cost schools money, but as CIO.com senior writer Clint Boulton reports in our September 2016 cover story, “Big Data on Campus,” analytics are helping universities and colleges identify and, when necessary, offer support to students who are at risk of falling behind — or even worse, leaving school without a degree.

As a parent with one daughter who recently graduated and another in her junior year, college is on my mind, as are the student loans and the job prospects that follow. Getting accepted to and thriving in the right college is a challenge every student and parent must contend with. The cost of a four-year degree continues to rise, and the stakes are high for students to “make it.” The only thing worse than six-figure college loan debt is that debt with no degree.

Colleges are taking note and turning to analytics to improve the odds, for their balance sheets and for their students. Boulton talked to CIOs and educational technology leaders at Purdue, Marist College and North Carolina State who are collecting, organizing and, more importantly, analyzing data to improve a national six-year graduation rate that hovers between 50 and 60, percent.

The cases we report on range from common sense to a bit creepy. For example, at Marist, CIO Bill Thirsk developed an analytics application that aggregates student GPAs and SAT scores, as well as demographic data, and correlates it all with information on how often students submit assignments and engage with instructors online. The application analyzes students’ browsing histories as they navigate the college’s systems.

The software tracks how fast students click on assignments, how fast they complete them and whether they chat with peers as they work on team projects. “We can very early on see which students are lagging,” Thirsk tells Boulton, saying the algorithms can anticipate whether students will earn a C or lower in the first two weeks of a class with an 85 percent rate of certainty.

If you see privacy concerns here, you’re not alone. With proper transparency and safeguards, however, a little creeping is something educators, parents and, yes, even students should be able to live with. Businesses need the workers, and students need to graduate — and college loans are onerous, even with a degree (I speak from experience).

Schools competing for students need to quantify their value. As Thirsk says, “If you don’t know what’s going on course by course then how can you say you have a great degree program?”