Signals: A Big Data Student Monitor Improves Graduation Rates at Purdue!

Dear Commons Community,

Since 2007, students enrolled in at least one class with Signals, a big data/learning analytics tracking program, saw a higher graduation rate than did students who were not in classes with the software, according to data from Purdue. Retention rates for those enrolled in classes with the tracking software were also higher than for those in classes without the technology. Students who had two or more Signals classes graduated within six years at a rate 21.48 percent higher than students who did not take Signals courses.  As reported by The Chronicle of Higher Education:

“Signals, a data-mining and analysis program, keeps track of how students approach class work. Taking in about 20 data points from Blackboard or other course-management systems, such as whether or not a student has completed online reading or watched online lectures, it measures the data against test and assignment grades, and “signals” students how they are doing with green, yellow, or red lights for each course. The signals are scheduled throughout each course by the instructor.

Matt Pistilli, a research scientist for teaching and learning technologies at Purdue, works specifically on student and faculty success. He says students have responded very well to the program.

“In this era of Big Brother, they aren’t concerned that we know all this information about them,” he says. “They want this to help them instead.”

By contract, he says, faculty members have been a little harder to work with. “Like Blackboard or a video conference or other pieces out there, this is just one more thing” they have to deal with, he says. Some faculty members also believe that students “who come to college should be self-motivated self-learners.”

With the software, a professor is able to “see the extent to which students have accessed materials, spent time engaged with the course, or viewed online lectures,” according to Mr. Pistilli. This is not done automatically but instead happens when an individual instructor logs in to see how students are doing.”

Anyone interested in learning more about big data and student monitoring, I published an article in JALN last year on the subject that includes references to Purdue’s Signals program.



Comments are closed.