Towards a Vision for Learning Analytics in Open-Ended, Student-Centered Engineering Program

Thursday, January 5, 2023
2:30 PM - 3:20 PM
Zoom

Presenters: Lauren Singelmann, Ph.D., Yuezhou Wang, Ph.D., and Elizabeth Pluskwik, Ph.D., Integrated Engineering

 

Minnesota State University, Mankato (MNSU), along with thousands of other educational institutions around the world, are committing to using data-driven analytics to better support student success. For example, predictive analytics tools such as Starfish have been shown to identify at-risk students and improve retention. However, these tools are largely designed for traditional educational systems where both student pathways and the data collected are greatly structured. In programs with more open-ended curriculum and self-directed learning opportunities, data collection and analysis lead to new unsolved challenges in the field of learning analytics -- the research field of using data to improve educational experiences.

An example of one of these open-ended programs is the Iron Range Engineering program in the Minnesota State University, Mankato College of Science, Engineering, and Technology. In addition to learning engineering fundamentals through coursework, students work full-time in internships and co-ops and get training in professionalism and design. The concept of Integrated Engineering presents many opportunities for students to design their own pathway towards a B.S. degree.

Although traditional learning analytics tools are not designed for such open-ended educational experiences, we argue that learning analytics will still be key to understanding student learning at a deeper level. To design and implement these tools, we propose using the community of practice theoretical framework. Communities of practice differ from traditional learning environments because they allow for self-organization and high levels of autonomy. These features support development of intrinsically motivated learning experiences, but they also lead to increased complexity in the classroom. However, common methods for predictive analytics including classification, clustering, and association analysis do not appropriately handle this complexity. Therefore, we propose using methods such as text mining and network analysis to understand the complexities of these learning environments rather than aim to simplify them.

This session will introduce the field of learning analytics, discuss the challenges of using learning analytics in open-ended, student-centered educational programs, share previous preliminary learning analytics work conducted at the Iron Range Engineering program, and present a framework for using learning analytics tools in these types of non-traditional programs in a way that is both equitable and empowering for students and instructors alike. Finally, we will share a vision for how this framework can be transferred to all learning environments to better understand the complexities of student learning.
"