Matt Bernacki (left) and Jeff Greene are exploring the value of data analytics in education, helping students learn effectively. (Photos courtesy of the UNC School of Education)
Two School of Education faculty are leading learning analytics research that uses data to efficiently deliver academic support to the students who need it.
Matt Bernacki and Jeff Greene are experts in the science of learning. They seek to understand how the human mind works and which methods are best for learning, recalling and applying information.
They’ve developed a learning analytics system that can be used in courses across the University to identify students who need additional support and connect them with the specific help they need, before their grades begin to suffer. Their system collects data from only those students who have consented to participate.
What does it look like to train college students on how to learn effectively?
GREENE: I always say that the science of learning is not rocket science, but it’s also not intuitive. Students have to be taught how to learn effectively.
One issue we often see in higher education is first-year students struggling as they adjust to learning in large 300-400 person introductory courses. Left on their own, many students will find it difficult to learn material in this new kind of class. They will likely still be using the same study methods they are familiar with from high school, which are often not the most effective methods, and many of them will then fail their first exam.
But using our learning analytics system, we can predict within the first two weeks of the semester whether or not students will be successful in a course. Our algorithm makes these predictions with 80% accuracy based solely on how students interact with course material, such as online textbooks. Then, for students who need it, we can intervene and offer support before their grades begin to suffer. This can look like prompting them to do certain exercises that teach new study methods, suggesting they pursue tutoring or attend their professor’s office hours, and more.
How does the use of learning analytics help deliver more effective academic support to students?
BERNACKI: We partner with course instructors who are committed to offering rich active learning environments. At the beginning of the semester, instructors let their students know that learning analytics can provide insights about learning that can improve the course. They also explain how sharing data can provide opportunities for feedback and support during the semester to students who choose to participate. However, sharing their data with us is completely voluntary for students.
For students who do opt in, we receive data about their learning behaviors, such as their engagement during online assignments; how they are interacting with online course materials that are designed to promote effective learning strategies; and how they are interacting with the tools that enable them to monitor their own learning.
These data inform prediction algorithms that help us identify who is learning well already, and who might benefit from feedback and support. To make this determination, the algorithm differentiates between students who are behaving like learners from previous semesters who have scored well versus those who have scored poorly in the course. Those who are on track to do well in the course will not be burdened with any unnecessary communication.
When learners’ earliest behaviors reflect similar struggles to what others encountered in previous semesters and they appear to need some support on “the science of learning to learn,” these materials are folded into their course work as an alternate resource in place of a more typical biology course activity. This way, students receive course credit for working on their learning skills, and are supported discretely to avoid any stigma, and without it requiring additional effort that cuts into their time for learning.
Delivering information in this way allows us to make learning support dynamic and adaptive to meet the students’ needs. These “Science of Learning to Learn” interventions have benefitted learners who complete them and improved achievement by as much as a whole letter grade for some students. Our learning analytics also provide valuable feedback to instructors after each semester is complete, letting them know which content and active learning processes have been most beneficial for learners in their courses.
Jeff Greene is the McMichael Distinguished Term Professor of Educational Psychology and the Learning Sciences. Matt Bernacki is an associate professor and the Kinnard White Faculty Scholar in Education. Distinguished and named professorships support renowned scholars and propel research at Carolina. These privately funded endowments help attract and retain the academic leaders of today, ensuring a state-of-the-art education for all Tar Heels.
As told to Audrey Smith ’10
Related Stories
