Abstract for: Understanding Projection Uncertainty with Stella

When calibrated models are used to project future outcomes, it is important to bound those projections with some degree of confidence. This hands-on workshop will introduce a practical, structured workflow that combines model calibration with sensitivity analysis to show the forecasting ability of a model. Designed for participants who want to deepen their applied modeling skills, the session will walk through a full end‑to‑end process including importing historical data, using Stella’s calibration functionality, assessing data conformance, finding approximate bounds on calibration parameters, and producing confidence bounds on future scenarios. The workshop will be done using a simple illustrative model so that participants can follow along and do the work themselves. The workshop will briefly compare this approach to Bayesian parametric sensitivity methods commonly used in advanced modeling contexts. We will discuss the strengths, limitations, and practical tradeoffs of each technique. Participants should bring a laptop; temporary access to Stella will be provided for use during the session. No prior expertise with calibration or sensitivity analysis is required, though familiarity with system dynamics modeling will be helpful. By the end of the workshop, participants will be able to apply a repeatable process for connecting historical data with future projections, understand how to incorporate evidence-based parameter constraints, and recognize scenarios in which these techniques provide the most value. This workflow will support stronger, more defensible modeling insights across a wide range of applications, from policy analysis to strategic planning and beyond.