Colleges and universities should look to Fitbit as a model for leveraging data to improve student learning outcomes, higher education leaders said at the ASU GSV Summit on education technology.
While wellness and education are two completely different industries, higher education leaders can take several lessons from Fitbit's use of data collection and predictive analytics:
1. Data tracking allows for real-time analysis and feedback
The University of Kentucky has incorporated data from all of its student systems into a high-speed analytical platform that allows the analytics and technologies team to access real-time streaming data.
Vince Kellen, senior vice provost of analytics and technologies at the university, said they are now considering how to use real-time data for in-class activity and feedback.
2. Data must be easily accessible for predictive analytics to be useful
Fitbit works because its platform makes it easy for users to quickly assess data such as their daily steps to gauge how close they are to reaching their goals. Higher education needs to follow suit.
"For these systems to work effectively, the data has to be in the right shape, and sadly, in postsecondary education, that's rarely the case," Renick said.
3. Behavioral data is essential
Many schools are only just beginning to take advantage of behavioral data. According to a recent survey, about 40% of higher education leaders say they have good quality data. Similarly, less than one-third of the respondents said they have both good data and the ability to use it.
But the right behavioral data can help administrators better understand how students' choices affect their outcomes—and "nudge" them to make better ones.
How one university used behavioral data to figure out where to focus their advising efforts—and retain an additional 400 students
4. Researchers should have freedom to play in a data 'sandbox'
Perry Samson, professor of atmospheric, oceanic and space sciences at the University of Michigan, combined research on student behavior with the school's student information system and learning management system. He identified certain characteristics of students most likely to succeed, but found it difficult to bring all of the disparate data into one place. Creating a "data sandbox" can make this process easier, Samson explained.
"I encourage universities to think about ways to take data, collect it from the LMS and SIS, put it all in one place, make a sandbox, and understand how learning and grades are related to student behaviors," he said.
5. Certain indicators can provide early warnings
If a Fitbit user wears a device but fails to meet his or her goal, some type of intervention is probably in order. The same goes for higher education.
Certain goals—like first-semester GPA—can tell us a lot about how students will perform over the long term and can flag those at risk of dropping out.
Nearly 40% of students don't open emails from their academic advisors. Are they opening yours?
6. Data might topple your long-held beliefs
Kellen discussed the ways in which the "massive transparency and democratization of data" prevents researchers from experiencing "data denial." It's hard to let go of beliefs we've long held to be true, but doing so is the only way to progress.
"Data will tell you things, and oftentimes it will tell you that your cherished belief is not true," Kellen said. "It takes time for administrators to realize that, sometimes."
7. People change their behavior when they can access their own data
Timothy Renick, vice provost and vice president for enrollment management and student success at Georgia State University, said he has seen a change in students after engaging with their own data.
"Once we got the students their own information, we found, en masse, that they changed their behavior," he said.
Learn how you can "nudge" students in the right direction
8. IT works better when silos are broken down
Even though they have historically been separated, IT and academic departments must work together to ensure that data is being accessed and used correctly (Devaney, eCampus News, 4/21).
Next in Today's Briefing
What the presidential election won't mean for higher ed