by Ethan Bernhardt, Associate Director of Reporting Services, Indiana University
Brent Drake of Purdue University offered an eye-opening presentation at the IACRAO Conference in October. Big data and machine learning are hot topics. Purdue is using these and other emergent technologies to both predict and improve student success. Students can opt into registering their devices (smart phone, laptop, tablet). Their movements can then be tracked by GPS while they are on campus. Their virtual actions when interacting with the device in addition to their times on campus, along with standard markers such as grades and attendance, can provide immediate feedback to each student, along with probabilities as to the likelihood of the student completing her/his degree program in a timely manner or at all. Studies indicate that it is not merely the feedback itself, but the timeliness of that feedback which ultimately can lead to a change in a person’s behavior.
Once the project was approved, the university computer services department at Purdue was able to develop and implement the infrastructure to maintain the 50 billion rows of data needed to support this initiative. Within eight months, students were able to reap the benefits. Through the use of compelling visuals, Brent showed what kinds of information students can see. In addition there are benefits for analysts who can see large chunks of the data and utilize statistical methods to show trends in student behavior. Tracking these trends over time can provide the evidence needed to construct models to further improve teaching methods, as well as inform those in administration as to the types of tracking of student performance which maximize graduation rates. Naturally, issues regarding privacy come into play here, but what researchers are discovering is that millennials are much quicker and more likely to adopt these practices than their older peers. Millennials have grown up in a world of the internet and social media. If they can see the benefit of having their movements tracked through their time on campus, they are very likely to opt in and provide this kind of information. Time will tell what other revelations lie in store for higher education utilizing big data.