By Kerrigan Stern


For The Diamondback

The University System of Maryland has implemented a system of predictive analytics to pinpoint students who are in danger of dropping out of school or losing their majors.

The system is “data being used to predict the future as opposed to simply describing the past,” said Ben Passmore, an assistant vice chancellor in the system. “What you have is a set of analytic tools that — instead of just telling you what has happened — it suggests from path patterns what will happen.”

The predictive analytic program analyzes past performance, the student’s course schedule and other factors when deciding on intervention. It also allows the students’ advisers to make a map of the graphical representation of grades received in a class with a certain professor.

On the graph, students who are 80 percent or more likely to succeed are shown in green, Passmore said. Those who are 15 percent likely to succeed are yellow, and the students who are 5 percent likely to succeed are red, he added.

With this information, a professor can decide where to focus his or her time inside and outside the classroom through efforts such as more office hours or a different curriculum, which may be directed at students of all achievement levels.

At other schools, the system is able to take other things into consideration, including dining points, to see if the student is an active member of the campus.

“It would be a good kind of early warning sign,” said Passmore, who said it is unknown whether the system truly uses dining points as a factor at this university. “Maybe they’re sick, maybe they’ve gone on vacation. Maybe a lot of things have happened, but the idea is … the entire experience of being present and around university campus, which is important for student success.”

But the program does have some drawbacks, including privacy and security of data, Passmore said.

“We know that some privacy experts have expressed concern in the U.S. that data could be used in a punitive way with students,” said Mike Lurie, the system’s media relations manager. “To the contrary, the data are useful in raising the chance that a student who is struggling in a certain academic track can be encouraged to seek help.”

Passmore compared the privacy issue to social media, which tends to take and disseminate users’ information for profit.

“We give away more information on Facebook, on Google, on Amazon, whose only purpose out there is to sell stuff,” he said. “But we feel sensitive about giving information to people … who are helping you help the student succeed.”

Student Government Association President Katherine Swanson also expressed concerns that predictive analytics could sway students away from their preferred major.

“I worry that they could be used to discourage students from trying their hardest to excel in their originally chosen major,” she said.

At the end of the day, it is the student’s responsibility to be successful, Passmore said.

“What you have is kind of a combination of the analytic system,” he said. “It’s the data team who puts this system together, the student service system … and the student themselves who have a responsibility to figure out what they can do.”