Teaching complexity in biology through agent-based simulations: the relationship between students’ knowledge of complex systems and metamodeling knowledge

Abstract

Real-world complex systems research seeks to understand how systems in the world can follow the same rules of complexity. Scientists have found similarities in processes—such as self-organization, micro-to macro-level emergence, and feedback loops—in seemingly disparate phenomena such as the spread of infectious diseases and how traffic patterns are formed. Our project, BioGraph 2.0, was developed to respond to the issue of students’ disjointed understanding of biology due to the fragmented nature of how high school biology is taught in high school classrooms. We hypothesized that by framing multiple biology concepts through the lens of complexity using dynamic simulations, or models featuring complex systems processes, students would be able to see complex systems as a unifying concept throughout biology. We built a series of units modeling phenomena on biological concepts such as gene regulation, ecology, and evolution using an agent-based modeling tool called StarLogo Nova. While previous research over the last decade of this project has highlighted students’ growth in complex systems understanding, in this study, we explored the relationship between complex systems and agent-based models. We investigated pre and post intervention data from over 300 high school students to determine how their metamodeling knowledge influenced their understanding of complex systems. Through a regression analysis, we demonstrate that growth in students’ modeling understanding significantly predicted growth in complex systems understanding. We further triangulate our findings with interview data from students who highlight the importance of the modeling tool to support their complex systems learning.