Colleges and universities have been under increasing pressure from President Obama to graduate students more quickly — and graduate more of them.
Higher education is looking to predictive analytics for part of the answer. Predictive analytics involves looking at past and current student data to predict future student performance. This is important because it can help students stay on track to graduate and alert them when they're falling behind.
Higher education leaders shared their thoughts on where they think predictive analytics are headed in conversations with the Center for Digital Education.
The problem with analyzing data is that it's typically stored in different locations and managed by different departments and organizations. But in the future, universities will take a holistic view of data and correlate grades, attendance, most recent course log-ins, and what was learned in previous courses, said Kari Barlow, the chief operating officer at ASU Online, part of Arizona State University. With that integrated data, the thinking is that universities would be able to individualize student instruction and ultimately increase retention and graduation rates.
"It's important that students are able to complete degrees in a timely fashion, both for them financially, and for the ability of the institution to best serve them and serve new students," Barlow said.
As more university systems consider predictive analytics, they'll likely figure out that it's something they should gravitate toward, said Jon Aull, executive director of operations for the Regents Online Campus Collaborative, an online effort governed by the Tennessee Board of Regents that includes universities, community colleges and technology centers. Data will also point out how and where universities need to improve in order to graduate more students.
"Predictive analytics helps us look under the rock and go, 'Ooh, we're having trouble in this area with these students, we need to fix it,"' Aull said.
At the course level, predictive analytics could help match students with appropriate programs, even down to the professors they should have. That's what the institutions in the Regents Online Campus Collaborative are piloting this year with a predictive analytics tool from Desire2Learn.
Predictive analytics also might test universities' tolerance for risk-taking. Eventually, universities will have to ask themselves, "What are we willing to invest in being wrong about?" said John Fritz, assistant vice president of instructional technology and new media at the University of Maryland Baltimore County.
"We're going to be getting to the stage sooner rather than later where it doesn't matter what you have access to or even what you think you're predicting," Fritz said. "What really matters is, 'What are you willing to act on?'"
Having common infrastructure also will be important, Fritz said. Universities have been creating their own custom analytics tools, including the University of Maryland, Baltimore County, which has now transitioned to Blackboard Analytics for Learn.
In the future, when similar tools and systems are used by all universities, they will have the data needed to figure out what interests they have in common.
"We have very different data," Fritz said. "But if the means to access and analyze that data becomes more standardized, then it becomes more possible to build on potential insights."