Â鶹ÊÓƵ

Synopsis

Introductory physics courses devote considerable time to teaching students how to solve well-defined problems: identify the relevant principle, set up the equations, and obtain an answer. These problems are an essential part of learning new concepts and developing fluency. But many of the problems students encounter later—whether in physics, engineering, or other technical fields—look very different. The relevant information may be incomplete, the appropriate model may not be obvious, and there may be no single prescribed path to a solution. What does it take to prepare students for this kind of problem solving? Research on expert problem solving suggests that much of the important work occurs before the calculation begins: experts make decisions about what matters, which physical models are appropriate, what assumptions are reasonable, and what information is needed. Context-rich problems provide an opportunity for students to practice making these decisions while still learning introductory physics. In this talk, I will describe my experience incorporating context-rich and ill-structured problems into introductory mechanics, electricity and magnetism, and related courses. I will discuss what students actually do when confronted with these problems, the challenges of providing appropriate scaffolding and assessing their work, and some of the unexpected difficulties and successes I have encountered. I will also consider how the emergence of generative AI changes the role of both routine and context-rich problems. Rather than simply making traditional problem solving obsolete, AI may provide an opportunity to reconsider what we want students to learn when we teach them to solve problems in the first place.