Student Success Is More Than a Numbers Game

While predictive analytics can help students succeed, we must not lose sight of the fact that college is also a time of exploration and self-discovery.

• By Andrew Barbour
• 12/06/12

With just 56 percent of college students earning degrees inside six years, I recognize that higher education is failing both students and employers. I also understand that predictive analytics can go a long way toward keeping kids in school who might otherwise flunk out. I admire such efforts, particularly in community colleges and career-oriented programs. However, I worry about unintended side effects at four-year schools.

Entering college, most traditional students don't have a clue what they want to do with their lives. College offers them a chance to try different subject areas, explore their interests, and--if they are taught properly--learn how to think. What's more, it's supposed to be hard. That's why I'm ambivalent about the kind of predictive systems that allow students to see what grade they might earn before they even choose a class (we profile such efforts in our feature "Digging for Gold"). In many ways, this runs counter to my own college experience, where intellectual breadth was prized. My university actually encouraged students to take classes that didn't fall into their academic wheelhouse--and at which they might not always excel.

By steering students toward courses where they are calculated to find most success, schools run the risk of creating a generation of narrow thinkers who attempt only what they're most suited to--and possibly never break a sweat along the way. And spare a thought for those demanding professors who set the bar high. Will their courses wither for lack of students?

Admittedly, most predictive systems focus on student performance after the semester begins. Even here, I worry about those student outliers--sometimes the most brilliant kids in the school--who don't fit into the mold and for whom the algorithm has no answer. So how do we throw a safety net under those students who might struggle, while still encouraging the kind of intellectual curiosity for which college is traditionally known?

For me, the answer lies in the human touch. As long as these predictive systems are paired with caring teachers, they have fabulous potential and can make a real difference. But faculty are key to success.

Which brings me to my second worry: MOOCs. With class sizes in the hundreds of thousands for some of these courses, the potential for meaningful faculty-student interaction is nil. If your fellow online learners don't provide you with adequate support, tough luck. Again, this may not be a problem for highly motivated students and postgraduates. But for young adults looking for guidance and assurance, it's a recipe for disaster.

As young adults seek their place in the world, they need to be pushed further than they think they can go, and they need teachers to motivate them to get there. No matter how massively the system is scaled, some things don't change.