Class Meeting Times
Lectures: 1 session/week, 3 hours/session
Prerequisites
15.873 System Dynamics for Business and Policy or equivalent, such as 15.871 Introduction to System Dynamics and 15.872 System Dynamics II.
Read the Introduction to the Course for Independent Learners for an in-depth explanation of how to utilize these online materials for your studies.
Software
Vensim DSS (Decision Support System). Vensim DSS is Ventana’s top-of-the-line software. It is most useful for building and analyzing differential equation simulations. Use the link to download Vensim: https://vensim.com/download/. There is a free personal learning edition (PLE) available for academic use.
Topics and Key Dates
| Class | Modeling Problem | Key Dates |
| 1 | Stocks: Horse market, oil market, and transportation system | |
| 2 | SIRSb with vaccination | |
| 3 | Energy efficiency co-flows | |
| 4 | Predator prey with carrying capacity | |
| 5 | Managing drug development pipeline | |
| 6 | Ponzi | Draft problem |
| 7 | Market growth with boom and bust | |
| 8 | Diffusion with endogenous scope and quality | Final problem |
| 9 | Dynamics of growth of a professional field | |
| 10–12 | Personal projects |
Readings and Preparations
Each week submit the model of the week, and read through the modeling assignment for the next week. Come prepared to discuss not only your work for the current week, but also key structural choices (e.g., aggregation level and stocks) and performance requirements (e.g., tests of intended rationality and expected dynamic patterns) for the next week’s problem.
Course Grading
This is a small graduate class focused on advanced skills and operates on high levels of mutual trust among participants and instructor, so grades are perhaps irrelevant. Yet to follow MIT guidelines, we will have grades for registered students, largely based on weekly model submissions and classroom participation. This class is self-paced in the sense that students will decide how much time they want to allocate to each modeling task. Primary deliverables are weekly model submissions and reports. The only other product you will produce is the personal problem definition, which we hope can be of use for future learners as well.
Students may (but are not encouraged to) skip up to two weeks of model submissions if they need a reduced workload due to exams or other personal constraints. If you plan to use that option in any week, you need to inform the instructor before the start of the class.
AI Policy
You can use generative AI tools if you find them helpful in your modeling process. Current AI methods do not offer much in terms of building more complex SD models, though they can sometimes give you the formulations for simpler and classical models. We do not expect polished reports, so the tools may not be relevant for editing, but occasionally useful in coming up with simple formulations or thinking through important mechanisms and tests. Please document your use of AI in your writeups and share any useful insights with the rest of the class, so we can all learn together.
