Skip to main
Binary code

New Spring 2022 CDS Course

System Dynamics

Instructor:  Dr. Ken Comer

Computational Social Sciences 655 in Spring semester 2022 will examine the use of system dynamics in the modeling of complex social systems for public policy, organizational planning, business strategy, and general forecasting.

System dynamics modeling for government, business, and industry is now in its seventh decade – it was the first and, so far, the most successful attempt to analyze the behavior of complex systems. This course will explore the many successes of system dynamics and teach a structured analytic method for applying this method to key decisions. There are numerous modern applications of SD techniques, including

  • The trajectory of epidemics and pandemics.
  • Economic forecasts for labor and housing markets.
  • Interaction of systems of systems with feedbacks for systems engineering.
  • The impact of policy changes and the evaluation and forecasting of public adaptation

This course will not let the math ‘get in the way’ and will provide students with analytic tools and skills in this important branch of modeling and simulation. The basic stock-and-flow concept has been matured, enhanced, and adapted to real-world processes so that it can be applied to a diverse portfolio of situations.

System dynamics practitioners have developed and matured excellent transferrable skills in the application of any quantitative technique to complex decisions. For example, through case studies we will explore:

  • The identification of all stakeholders
  • The elicitation of problem statements and structure from stakeholders
  • The need for explicit definition of model assumptions and model boundaries
  • The balance of model simplicity vs. model accuracy
  • The overpowering and potentially misleading nature of graphical model depiction
  • The process of constantly refining a model based on data, output, and client feedback
  • The psychology of model cognition and the uneven recognition of system behaviors (i.e., negative feedback has been shown to be overlooked)
  • The use of dynamic models to uncover key decision elements that are not evident by ‘data analytics’ alone
  • The benefits and risks associated with fitting a model to data
  • The cross-domain opportunities that exist by applying mature models in a different context (models of epidemiology, for example, work well for understanding innovation)

In addition to developing these important individual concepts, students will also experience the application of a mature body of practice to unstructured and poorly defined problems. This will illustrate the tradeoffs and contradictions inherent in the modeling of complex systems.

The course will be based on real-world case studies in which system dynamics modeling techniques were applied to policy or business decisions. Fortunately, the enduring partnership of system dynamics with government and industry has produced a rich literature, which includes both successes and failures. Students will be able to build quantitative analytic ‘process’ skills – such as elicitation techniques – that are useful outside the system dynamics domain, in areas such as big data analytics, agent-based modeling, business process modelling, and discrete-event simulation.

Instructor: Ken Comer, whose PhD is in Systems Engineering Operations Research, and who has 35 years’ experience applying quantitative analysis, modeling and simulation to difficult real-world problems. Dr. Comer currently serves as the Lead Mission Analyst for the Office of Undersecretary of Defense for Research and Engineering.