How do colleges predict which students might drop out? The largest university in America has found an answer in big data.
The University of Central Florida in Orlando has 66,000 students, and low-income and first-generation students who are at risk for quitting could easily get lost in the shuffle. But UCF wants to make sure they don't. By using predictive analytics, the university identifies at-risk students and intervenes before they drop out.
"This is a new way to really find those students who need our attention the most," UCF vice president Maribeth Ehasz tells Scott Pelley in the video above.
Ehasz has looked through 20 years of data to find out what students who have left school have most in common. Among many categories, two results surprised her: avoiding the gym and living in Volusia Hall, a dorm that had historically been set aside for last-minute applications. Today, if an underprivileged student lives in Volusia and doesn't work out, counselors step in to help the student succeed.
Ehasz says the approach seems to be working. "Over the past seven or eight years since I've been tracking, I've seen that our retention rate has increased, particularly for the first-gen students, by about 10 percentage points," she says. "That's pretty big."
UCF is a partner school in the University Innovation Alliance, a coalition of public research universities that work to increase the number of low-income students who graduate college. In four years, Alliance schools have increased low-income student graduation by nearly 30 percent.
The UIA receives funding, in part, from the Gates Foundation, whose founders are trying to narrow the country's wealth gap through education. Paying the tuition and expenses for more than 20,000 students across the country, Bill and Melinda Gates have spent more than $1 billion dollars to help low-income students attend college.
To watch Pelley's 60 Minutes piece on the Gates' effort to close the wealth gap, click here.