IC Researchers Utilizing OMSCS as Test Bed for Wearable Tech in Online Learning

Sep 6, 2018

A team of researchers in the School of Interactive Computing (IC) and the Online Master of Science in Computer Science (OMSCS) program will investigate the feasibility of using wearable technologies and other types of sensing data to provide context in online learning environments.

This is the first funded project that uses OMSCS as a test bed.

Under a $300,000 grant from the National Science Foundation, IC Assistant Professor Lauren Wilcox, along with co-principal investigators Betsy DiSalvo (IC), Thomas Ploetz (IC), and David Joyner (OMSCS), will set up an infrastructure for using wearable technology and interaction analytics to capture students’ experiences with online courses. They will also investigate which personal interactive computing technologies are effective in capturing and modeling context, and what correlations exist between wearable data, analytics from online behavior, self-reports of stress and anxiety and learning outcomes. The initial aim is to determine if the use of wearable technologies could better inform online course delivery, and improve retention and learning outcomes.

“When we are instructing online courses, we lose an important view into the student experience,” Wilcox said. “Some students might be paying attention while others aren’t. Some might be paying attention, but they still aren’t learning. We want to better understand these scenarios and use our knowledge of them to inform better online learning experiences.”

“Analyzing data from wearables worn by students during online course instruction could help us understand and recognize these scenarios,” Ploetz added.

Such a study may not be possible at the vast majority of research institutions, but the presence of the OMSCS program at Georgia Tech affords a unique opportunity. Because of the high volume of participation and enrollment, not to mention the number of quality dedicated professors, the researchers expect to establish reliable conclusions and create a foundation for future research

“We think the OMSCS program is a strong test bed for this research because students are motivated to succeed and course completion rates are very high, and yet the course content and assessments remain extremely rigorous,” said Joyner, the associate director of student experience in the College of Computing and a longtime lecturer for OMSCS.

“These students experience the same stress, engagement, discouragement, and triumph as traditional students, but online instructors cannot see these states. Wearable technologies may help identify when these states occur and whether they correlate to desirable learning outcomes.”

The project is a two-year study during which the researchers will establish ground truth on student success and satisfaction. Are they generally happy? Do they disengage or pay greater attention to a specific lecture? What events trigger indicators of stress or anxiety, and at what point is that detrimental to learning?

“The idea is not to provide data on individuals to instructors,” Wilcox said, addressing concerns of student privacy. “First, we hope to see whether we can collect these data points, understand what they might mean for learning, and then provide anonymous aggregated feedback to instructors. It’s also about how we can help adapt these learning experiences to the individual students.”

Down the road, it could also be an important test bed for things like test anxiety or understanding what a flow state – or, colloquially, being “in the zone” – looks like and what features of a lecture lend themselves to it.

“One of the most exciting aspects of deploying this type of research in OMSCS is the potential scale for future research,” DiSalvo said. “This grant is laying the groundwork for future research on designing learning, building upon theories of learning with design-based research both at a scale and with detail of individual behavior and feedback that we have not had access to in the past.”

While the initial applications of this approach are in online courses, access to this type of data could be used to design many new learning environments.

“With just-in-time feedback to the students, we could provide customized learning that really moves away from the traditional class structure,” DiSalvo said.

Beyond learning, with more and more aspects of daily life going online, Wilcox said that she could also see implications of the findings from this study influencing the design of other online environments, such as job training.

“The ultimate goal is to create online learning environments that promote positive human interactions and consider human health and wellness an integral part of the design,” she said.