Understanding learning through data

A fresh way of developing new educational tools

CEHD’s new Learning Informatics Lab lies at the crossroads of computation and interdisciplinary collaboration. Learning informatics is an emerging field that transforms massive amounts of educational data into tangible information educators can use to develop innovative learning methods. CEHD’s new lab is designed to bring together researchers from across disciplines to use learning informatics to find solutions to some of the most challenging issues in education today.

As an illustration, lab co-director Bodong Chen cites the research a doctoral student is currently conducting in an urban Twin Cities school. The researcher surveyed a student cohort about peers, teachers, and others in their lives who support them. The gathered data was then used to create a social network map for the teachers to talk about with each other. Using learning informatics, the researcher was able to provide teachers with different perspectives on how best to nurture students, such as through a caring peer or role model from their cultural community. These new perspectives will foster the support students need to succeed.  

“Looking at test scores is not enough,” says Chen, an associate professor of curriculum and instruction and Bonnie Westby Huebner Chair in education and technology. “We need to look at the whole social, emotional, and behavioral picture [and] examine the data to find new patterns. Then we can put that in the hands of educators and make real-world changes. That’s the most exciting thing to me.”

Interdisciplinary by design, the lab opens doors for people across campus to pursue their shared interest in using computation to improve education. They bring with them a range of expertise, including learning sciences, information science, educational technology, and data mining, says lab co-director Sashank Varma, Bonnie Westby Huebner chair and professor of educational psychology.

“We will generate research we wouldn’t have been able to do before if we just stayed in our departments. Now we will get people together in the same room, conduct research in classrooms and in the lab, use machine learning and statistical analysis, and ultimately produce new insights that will improve education,” Varma adds. “We will train researchers who will be on the leading edge, with one foot in the educational world and one foot in the informatics world.”

Learning informatics doesn’t necessarily involve finding new sources of data—it strives to develop fresh ways to look at the reams of data that schools, educational systems, and governments already collect. The data can come from anywhere. One of Chen’s projects has elementary students post their ideas in an online forum and build on each other’s ideas to make them better.

“They talk about whatever that is of interest to the class,” he says. “For example, some classes talk about why leaves change color, how light travels, what is soil, how worms sense light, how birds fly, why North America is having the highest CO2 emission per capita, and so on.”

Chen, along with an international coalition of researchers, uses the student-generated discussion data to develop learning analytics tools. The tools aim to help teachers monitor social interaction in a class, map students’ discussions with the curriculum, and show students the big picture of their ideas—thus helping them to make informed decisions about their own work.

“There is a natural need for tools and techniques for understanding learning processes,” Chen says. “Part of that motivated people to apply data science methods to understanding learning and use computation to augment their research. Our college is doing very exciting work with the Learning Informatics Lab to create new interdisciplinary space for achieving more ambitious goals.”

For Keisha Varma, an associate professor of educational psychology and the University’s vice provost for equity and diversity, being a lab member will help her integrate more data collection and analysis in her research. She envisions using learning informatics to broaden her work in effective science education by assessing students’ interactions with their learning environment.

She plans to collaborate with lab members on designing learning environments that reveal what and how students are learning. “It’s giving another dimension of information,” she says. “You can collect information about students’ behavior and interactions in the learning environment. You’re seeing more of their dynamic behaviors and not just a static output of data.”

Learning informatics strives to develop fresh ways to look at the reams of data that schools and educational systems already collect.

The lab’s confluence of experts also will attract prominent speakers to share their expertise and industry professionals who might seek educational research partners to advance their technology efforts. Eventually, the lab hopes to offer a graduate certificate in learning informatics, Sashank Varma says.  

As computation continues to evolve into a foundational aspect of higher education, CEHD’s Learning Informatics Lab will be leading the way through its mix of data crunching and collaboration.

“Computation gives the ability to crunch through huge amounts of data and see the learning trajectory for kids, and also to identify what stops their progress,” he says. “People across the U are interested in the same problem, just different parts of it, and the new lab will bring them together.”

Story by Suzy Frisch | Photos by Greg Helgeson | Spring/Summer 2020

More information here.