Editorial: USF uses big data well to help students succeed

In a university teeming with 30,000 undergraduates, where the average freshman's GPA exceeds 4.0, it would be easy for struggling students to get lost and drop out. But the University of South Florida, using sophisticated software, is making a concerted push to identify students who are having trouble and quickly intervene to help. The results are unequivocal: a narrowing achievement gap, higher retention rates and more students graduating on time.

The Tampa Bay Times' Claire McNeill reported on the data-driven effort, which has helped dispel many assumptions about what causes students to succeed or fail. USF contracts with Civitas Learning, whose software can track when assignments are turned in late, monitor exam scores and notice if a student suddenly stops engaging on class discussion boards. Those red flags trigger intervention. A financial counselor, professor or resident adviser can be dispatched to check in with the student, find out where the trouble lies and connect him or her with services such as tutoring, counseling or financial aid. It works from the curriculum side as well. USF can use the software to analyze which courses are strong graduation predictors, McNeill reported, and devote extra resources to ensure students master those make-or-break subjects.

USF's focus on intervention and improving its graduation rate predated Civitas. It started with freshman surveys that asked students to identify their needs. The university added advisers and other programs to help students feel engaged on the notoriously commuter-heavy Tampa campus, and there were some gains. When the state began tightly tying funding to meeting metrics such as student retention, that led to the $286,000 Civitas contract. It is clearly money well spent. USF's six-year graduation rate in 2016 was 66.3 percent, better than the average across state universities and 14 points higher than five years ago. Its achievement gap — the difference in graduation rates between white and African-American students — is just 2.1 percent. Nationally, the spread is more than 20 percentage points and notoriously difficult to close. Black students' six-year graduation rate at USF has increased to 69 percent. Those statistics are the reward for USF's concerted dedication to helping all students succeed, and it's an approach that should be emulated.

Ethics experts have raised some legitimate concerns about the privacy implications of tracking students with real-time software. These powerful analytics tools must be used only to support students, not as a means of weeding out underachievers. For example, poor grades in particular courses should not prompt advisers to steer students toward different tracks in order to boost a university's bottom line. And until more is known about how effectively the system works over time, it should continue to function in the background. In order to prevent biases from "creeping into the teaching and grading process," a school vice president said, USF has so far held off on handing over the software to faculty. That prudent approach still encourages faculty to connect with students personally and be proactive when they start to fall behind.

In the still-new (and somewhat unnerving) era of big data tracking, USF has found a wholly useful and constructive application. By keeping close, real-time track of student performance, USF is spotting potential problems in time to correct them — and the school and its students are reaping the benefits.