Data analytics has received a lot of attention in higher education in recent years for its power to improve retention. Schools across the country are exploring ways to find patterns in things like student demographics, campus activities, and course participation to develop early warning systems that alert counselors and the students themselves that they might be off-track. For most institutions, it is not a matter of collecting new data, but of merging datasets and looking for the patterns.
Indiana’s community college system, Ivy Tech, is one of the institutions pushing the boundaries of traditional data analytics. And it is finding patterns relating to a lot more than student performance. Lige Hensley, Ivy Tech’s chief technology officer, says the college produces about 100 million rows of student-related data per day. Like many colleges, its first step with data analytics was to create an early warning system to encourage proactive conversations with students in danger of failing their courses. But beyond student performance, Ivy Tech has turned its analysis on faculty.
Analyzing similar student and course data can identify faculty with highly irregular outcomes. Hensley said one review of the data caught four faculty members with patterns that were “drastically different from other members,” which gave the school's vice president of online programs information for employment decisions.
Machine learning algorithms have helped staff members see things they couldn’t before.
“We accidentally stumbled across some patterns for financial fraud,” Hensley said. The patterns were clear but not clear enough to catch the eye of a human technician. They needed to be flagged by more sophisticated programming. Now, Ivy Tech incorporates a fraud prediction algorithm in its financial aid process.
The college has also discovered how to improve its online courses by analyzing the length of time students spend on certain modules. A quick data analysis can highlight content that should be made clearer so students don’t get stuck. Sections of courses have actually been redesigned, thanks to this analysis.
Hensley said he has a “a lot of pie-in-the-sky” applications for the data analysis tools. One use that is in a prototype phase deals with student support. The college has natural language analysis tools that can track what students are saying on social media platforms and has the ability to look for keywords. If students are behind on school payments and mention money issues on their social media profiles, for example, Ivy Tech employees could be prompted to reach out with advice about financial counseling opportunities or carpooling. Just like mentioning a product in a post consistently triggers targeted advertising, Ivy Tech hopes to use existing data to better help students.
Hensley said Ivy Tech leaders have had many conversations about potential privacy concerns based on these innovative applications. But in many cases, the new uses don’t require collecting any new data, they just require making use of it. Besides, the intentions are pure, Hensley said.
“We’re not doing this to be Big Brother,” Hensley said, “we’re doing it to look for trends and assist students.”
With a recent transition to Amazon Redshift and its nearly limitless capacity to hold data, Hensley and Ivy Tech are on the cusp of making data analytics a fundamental part of operations.
Would you like to see more education news like this in your inbox on a daily basis? Subscribe to our Education Dive email newsletter! You may also want to read Education Dive's look at whether smartphones are making campus blue lights obsolete.