Big Data's Arrival

New students are more likely to drop out of online colleges if they take full courseloads than if they enroll part time, according to findings from a research project that is challenging conventional wisdom about student success.

But perhaps more important than that potentially game-changing nugget, researchers said, is how the project has chipped away at skepticism in higher education about the power of “big data.”

Researchers have created a database that measures 33 variables for the online coursework of 640,000 students – a whopping 3 million course-level records. While the work is far from complete, the variables help track student performance and retention across a broad range of demographic factors. The data can show what works at a specific type of institution, and what doesn’t.

That sort of predictive analytics has long been embraced by corporations, but not so much by the academy.

The ongoing data-mining effort, which was kicked off last year with a $1 million grant from the Bill and Melinda Gates Foundation, is being led by WCET, the WICHE Cooperative for Educational Technologies.

A broad range of institutions (see factbox) are participating. Six major for-profits, research universities and community colleges -- the sort of group that doesn’t always play nice -- are sharing the vault of information and tips on how to put the data to work.

“Having the University of Phoenix and American Public University, it’s huge,” said Dan Huston, coordinator of strategic systems at Rio Salado College, a participant.

According to early findings from the research, at-risk students do better if they ease into online education with a small number of courses, which flies in the face of widely-held belief in the benefits of full student immersion.

“Each of the different institutions has a very different organizational structure for how they deliver courses,” said Sebastián Díaz, the project's senior statistician and an associate professor of technology at West Virginia University. “What the data seem to suggest, however, is that for students who seem to have a high propensity of dropping out of an online course-based program, the fewer courses they take initially, the better-off they are.”

That discovery warrants a rethinking of how to introduce students to college-level work, the researchers said. And the problem of too many concurrent courses may be worse for students who depend on financial aid.

Students can only receive the maximum Pell Grant award when they take 12 credit hours, which "forces people into concurrency,” said Phil Ice, vice president of research and development for the American Public University System and the project’s lead investigator. “So the question becomes, is the current federal financial aid structure actually setting these individuals up for failure?” (This paragraph has been modified because of a factual error.)

Early Warning System

Most of the project’s participants were already collecting sophisticated data about their students. But researchers said this research is on a different scale, as are its practical applications.

The downside, however, is that the data sets are so detailed that analyzing them is far from over. “It’s going to be a taxing process,” Ice said.

While results from the project, which is dubbed the Predictive Analytics Reporting Framework, are preliminary, participants are already putting them to use.

Rio Salado, for example, has used the database to create a student performance tracking system.

The two-year college, which is based in Arizona, has a particularly strong online presence for a community college – 43,000 of its students are enrolled in online programs. The new tracking system allows instructors to see a red, yellow or green light for each student’s performance. And students can see their own tracking lights.

On January 9 Rio Salado turned on the switch for the system across 80 percent of its online courses, according to college officials. It measures student engagement through their Web interactions, how often they look at textbooks and whether they respond to feedback from instructors, all in addition to their performance on coursework.

Michael Cottam, Rio Salado’s associate dean of instruction, said the college would not have been able to track students with the same detail and real-time speed without the research from the WCET project.

“Previous to now, we haven’t had the data,” Cottam said. “This gives you something tangible, something real.”

Match.com for Higher Ed?

The data set has the potential to give institutions sophisticated information about small subsets of students – such as which academic programs are best suited for a 25-year-old male Latino with strength in mathematics, for example. The tool could even become a sort of Match.com for students and online universities, Ice said.

That application is nowhere near to being a reality, in part because institutions are loath to share competitive information with each other, or the general public. But researchers said the project will almost certainly help other colleges follow Rio Salado’s lead in using predictive analytics to help design better academic programs.

“If institutions of higher education did more of this type of analytics," Díaz said, they could tell their prospective students: "Look, these are the kinds of students who tend to have more success at our institution.”

The project appears to have built support in higher education for the broader use of Wall Street-style slicing and dicing of data. Colleges have resisted those practices in the past, perhaps because some educators have viewed “data snooping” warily. That may be changing, observers said, as the project is showing that big data isn’t just good for hedge funds.

And the researchers have already achieved one of their primary goals, which was to prove they could create such a large, workable database. In addition to studying the data, the project’s leaders hope to begin a second round soon, maybe adding up to 18 new institutions.

What comes next depends on how colleges use what they learn. Going public with hard facts gleaned from such a database, Díaz said, could help students and their parents make better decisions. “Rather than just going on rankings done by a particular news agency,” he said, they could, “really look at tailoring which institution provides the best fit for a particular individual student.”

Steve Kolowich contributed reporting.