Lecture F3 (2022-04-12): Chapter 10, Model Validity, Mental Models, and Learning (Morecroft, 2015)

In this lecture, we review the key points of Chapter 10 from Morecroft (2015), with some additional connections to literature from Frank Keil, George E.P. Box, and a few others. The chapter focuses reviews the purpose of models that fall all over the modeling spectrum -- from realistic, analog models to less realistic (but highly generalizable), simplistic, metaphorical models. We discuss how the process of building models (even simple models) helps us "transition" our mental models to more sophisticated and deeper levels of understanding and move ourselves away from the "illusion of depth" (or "shallows of explanation") that we might have before forming such models/formal theories. We extend this idea to using models to help achieve shared understanding with other experts whose expertise might differ from our own. We then pivot to discussing how we build confidence in the formal models we build -- ensuring that they have the right boundaries, structures, and equations and that they produce the right behaviors and even allows us to learn about the original system through experimenting with the modeled system.