Big Data on the College Campus

• By Bridget McCrea
• 10/10/12

The information age has been both a blessing and a curse for institutions of higher education. They have suddenly found themselves the purveyors of larger and larger data sets which are also known simply as "big data." Having reams of data to sift through and use can be a good thing when it comes to pinpointing problems and making decisions, but the distillation process itself takes time, effort, technology, and human resources.

The educational sector harbors one of the largest big data repositories. According to McKinsey Global Institute’s most recent numbers, the sector had a total of 269 petabytes (one petabyte = one quadrillion bytes) of data under its domain in 2009. That makes education one of the U.S. economy’s top 10 sectors storing the largest amount of data.

Information Overload

Big data isn’t going away. The world will generate 1.8 zettabytes of data this year alone, according to IDC's 2011 Digital Universe survey, and data volume worldwide increases by at least 59 percent annually, according to Gartner. These increases will further challenge college IT departments to not only find ways to capture and store useful data, but also provide constituents with fast, reliable, useful information.

At Arizona State University’s Advanced Computing Center, Director Frank Timmes and his team are attempting to strike that balance. "We’ve embraced big data on campus both from the business analytics standpoint and in terms of our internal IT processes," said Timmes. "Those are the two sides of the big data coin for universities."

On the business analytics side, for example, ASU taps its vast databases to manage retention rates and dropout rates and to better tailor the student experience on campus. With 72,000 students enrolled in more than 250 majors, ASU’s eAdvisor system allows incoming freshmen to select their majors and then map out their key courses. Failure to sign up for those courses sends up an "off track" red flag for advisors, who can proactively intervene.

Class planning and the hiring of instructors are easier when big data comes into the picture, said Timmes. If student enrollment increases in a certain major, for example, ASU can scale up in advance to meet that demand, rather than waiting until problems begin to surface. "We can spot trends pretty quickly," said Timmes, "and scale up or down easily when we have that kind of information at our fingertips."

Prime the Faculty

At Temple U rollout of lecture capture was positioned as a test. The university rolled it out with a group of beta users. "We got them to be the spokespersons for capture. Not me," DeAngelo says. "They want to hear other faculty talking about its value."

So that's what he expected at UCSF too, but he was pleasantly surprised. "There was a readiness factor here. It's like, people were waiting for this. They all jumped on. I could not have foreseen the rapidity that this caught on here. I was prepared to have to do a lot more selling." For example, whereas faculty at Temple had questions about intellectual property--who owns the lecture recording and can the whole world see it?--"here it just was not."

DeAngelo's approach in either kind of atmosphere is to get buy-in from everyone and let people know what's happening "every step of the way." He's also learned over the years to not present anything as a fait accompli. "I present everything as a pilot... We're all adventurers together when you do a pilot. And you're not saying, 'It's done, it's over.' You're saying, 'We're going to learn together.'"

Small Scale, Big Data

At Yale University’s Digital Media Center for the Arts, Lee Faulkner, media director, said managing big data has been a key concern for years. The center’s 1,000 music, filmmaking, and photography students, who store their artistic creations on a dedicated server, generate much of the data. Managing the memory-hogging data is an ongoing issue for the center, which is now working with the university’s IT department to roll out a campus-wide, shared-storage solution.

"We’ve been dealing with video and graphics files for years and they just keep getting bigger and bigger," said Faulkner. "Luckily, the technology has also improved and allowed us to accommodate, store, and access this data in a pretty efficient manner." But Faulkner wants more. He envisions a time when the large quantities of information generated by the Center are stored in a hosted format online and no longer under his department’s domain.

"That will allow us to reduce the size of the files, speed up delivery, and be much freer," said Faulkner. It will also help the Center retain more of its students’ work – an exercise that’s cost prohibitive in its environment. "The data we’re working with is fairly temporary. We can’t keep it forever – we’re not Google," said Faulkner. "We can’t just have a pot that keeps expanding and expanding because technology is still pretty expensive on a per-terabyte basis."

The Bottom Line

As universities nationwide seek out effective ways to harness and utilize big data, many are turning to technology to help them achieve those goals. At ASU, Timmes sees analytics, or the discovery and communication of meaningful patterns in data, as a saving grace for universities looking to get their arms around their growing databases. "Most colleges already have a few people poring over data and extracting useful information anyway," he said. "It’s really just a matter of adjusting those skill sets and adding computing and an analytic engine to the mix."

When that "mix" is in place, big data becomes less of a burden and more of a strategic asset that colleges can use to make smart, fast decisions across many different functions and departments. Institutions can analyze big data around student enrollment and progress. They can then use the information to retain a higher number of incoming freshmen, improve the overall number of degrees awarded, and/or attract new students.

All of those improvements can result in revenue increases for the institution. "We’re not just talking about abstract knowledge here," said Timmes. "At the end of the day, being able to harness and use big data actually impacts a campus’ bottom line."