Work In Progress: Development of a Simplistic Agent-based Model to Simulate Team Progress Within an Innovation-based Learning Course

This work in progress discusses the development of an agent-based model used to simulate probabilistic team behavior within an Innovation-Based Learning (IBL) course. Innovation-Based Learning is a non-traditional learning model that encourages students to learn and solve complex problems relevant to current societal challenges as defined by local and federal funding opportunities. IBL pushes students to think innovatively, especially on problems with "unknown unknowns" typical of complex systems. Our research team has implemented the IBL learning model into a cardiovascular engineering course currently offered within four institutions across the United States. When entering the class students are encouraged to define their own learning. They must identify a current relevant problem, form diverse multidisciplinary teams across the multiple institutions involved, and then work collaboratively towards an innovative solution. Throughout the course students are expected to apply the core course concepts they learn into their innovative solutions and track all learning objectives on the course's online platform. In previous semesters instructors have noticed that many students flourish and perform exceedingly well within the IBL model, but a few students seem to struggle consistently. Understanding the components of what makes teams successful versus unsuccessful is crucial for enhancing the IBL model and scaling it to larger student populations. The complex nature of students, along with the various causes and effects that take place within team dynamics geared our research team towards modeling the IBL course within NetLogo, a multi-agent modeling software. NetLogo's modeling software allows complex systems to be broken down into key components or inputs controlled by simple rules that simulate cause-and-effect interactions of various agents and lets researchers assess the agents' effects on the system as a whole. Creating a model capable of simulating student and group interactions means researchers can introduce various factors into the model and assess how the group dynamics alter as a result. Such results would provide insight into what factors are beneficial or detrimental for group progress. The authors have identified several key components considered influential on student success in the IBL course: group size, time devotion, project connection, team citizenship, and learning outcomes (defined by Webb's Depth of Knowledge and an Innovative Impact Scale). These inputs were randomly assigned to agents and agent-sets within the modeling system. Simulations were then performed, data collected, and team progress assessed. To determine the model's accuracy, core instructors were asked to define these key components for a previous cohort of IBL students. The defined inputs by the instructors were compared and accorded prior to being integrated into the model to remove randomization. The results of team progress in the interpolated model were then analyzed and compared to actual student teams' progressions within the course. Any results of the models are probabilistic due to the nature and complexity of agent-based modeling. However, the insights into what key factors suggest successful team progress are influential in further advancing a course that harnesses educating students in complex innovative learning environments.