One morning last summer, Zulmaly Ramirez, an academic advocate who advises undergraduates at the University of South Florida, logged on to her computer and saw a notification that a new freshman was at risk of dropping out. The student, an off-campus commuter, hadn’t been signing in to the course portal, where reading assignments are posted, and his grades were slipping, the software showed. Ramirez asked the young man to stop by.
In person, the teen confirmed exactly what the computer program’s algorithm had predicted. His half-hour drive to campus made him feel removed from the other students, he had yet to decide on a major, and he had recently broken up with his girlfriend. Ramirez proposed some quick fixes. She introduced him to the ultimate Frisbee team, helped him settle on a business track and personally walked him to the counseling center to make his first appointment.
Ramirez’s intervention can be credited to Civitas Learning, a software company that sorts reams of student data to warn counselors, in real time, which students are in the greatest danger of dropping out — before the semester has ended and grades have been posted. The company, based in Austin, Texas, also has programs designed to help students pick classes and allow administrators to track what impact they have on student performance.
Civitas, which has contracts with Texas A&M, the University of Arizona, Penn State University, Morehouse College and hundreds of others, has pledged to boost graduation rates by 1 million more students each year, before 2025. (Economists predict America must add up to 23 million skilled college grads to its increasingly tech-centric workforce, by 2025, to be globally competitive.) The company plans to reach that goal by completely revamping the function of advisers in higher education.
“Most students’ relationship with their adviser is fairly transactional. ‘What are the classes that I have to take next?’ And, ‘How do I enroll?’ Unfortunately, the conversation is hurried and infrequent,” says Charles Thornburgh, one of Civitas’s two co-founders. “Hopefully in the future, more tools will provide more personalized recommendations to students, with both the student and adviser coming in dramatically better informed about where the student is on the journey to success.”
Previously, most college advising departments merely guessed who might not graduate on time. These counselors often based their speculation on whether a student was meeting traditional markers of success, like a high grade point average — a policy backed up by intuition, not evidence. Civitas, by contrast, starts with a review of a college’s historical data to detect which factors recur among dropouts, a more accurate way to develop a school-specific predictor.
Often, the results of this analysis surprise even veteran administrators. One of Civitas’s recent findings, for example, showed that GPAs were nearly meaningless when correlated with retention rates. A student with a 2.0 was no more likely to quit than a high achiever with a 4.0. Rather, the surest sign a kid wouldn’t make it was his grade in a freshman writing course.
Using Civitas, an administrator can easily see which students are thinking of quitting. They can also test how well an intervention can reverse a downward trend. Armed with a vast archive of historical data, Civitas’s software first digs up the records of past students with similar circumstances. Then, it analyzes how intervening would change the students’ learning trajectory, compared to the past. “There’s an opportunity there for educators and administrators, who’ve been operating in the dark forever,” he adds. “They can become more scientific.”
Of course, there’s a danger in placing too much faith in numbers. In the wrong hands, predictive analytics in education might divert resources (or deny college admission) from students who are careening toward failure anyway. But Civitas maintains that its approach is intended to direct help to those students who need it most, not to take it away from their classmates. Thornburgh notes that the education system already relies on an insidious predictive model. Fixed characteristics like family wealth, race and gender are seen as factors in student success — inherent conditions that, Thornburgh points out, can’t be changed or reversed. Luckily, as he’s found in his research, demographics aren’t the best way to predict who stays in school. “How students engage while on campus is dramatically more important than anything else, and that’s what really drives our model.”
Civitas emphasizes its role as a tool to support more personal academic advising. After the software flags a student, the intervention comes from a counselor, not a machine. “With our freshmen, even though they do use their phones and technology a lot, I’m always surprised by how much they enjoy just sitting down for 30 minutes or an hour,” says Ramirez. “I see students change dramatically when they have a meeting face-to-face, rather than receive alerts on their phones.” Especially when isolation drives disengagement, that human interaction can go a long way.
So far, it seems to be working. At the University of South Florida, retention rates that had plateaued for years finally surpassed 90 percent this year, reports Paul Dosal, vice provost for student success. And while the data’s not in yet, he expects that USF will finally crack a 70 percent graduation rate very soon, a huge step as the college looks to boost its prominence.
Across academia, researchers spend plenty of time conducting research in the humanities, sociology and the hard sciences. But they too rarely turn that critical eye to assessing the best way to teach the degree-seekers in their own lecture halls. With Civitas, these professors and administrators can begin to study themselves.
“There’s a lot of capital, time and energy spent on educating students, and we should find a way to make sure that we keep getting better at it,” Thornburgh says. “As educators, we have to learn from each other and every student’s journey.”
This article is part of the What’s Possible series produced by NationSwell and Comcast NBCUniversal, which shines a light on changemakers who are creating opportunities to help people and communities thrive in a 21st-century world. These social entrepreneurs and their future-forward ideas represent what’s possible when people come together to create solutions that connect, educate and empower others and move America forward.
Let’s fix this country together.