Abstract for: Bringing Data Into Dynamic Models: Guidelines for Advanced Estimation Methods

Dynamic, non-linear models require the development of customized methods for formal estimation. With the increasing availability of data and computational power, opportunities for advancements in these methods are abundant, though many audiences remain unfamiliar with their application. In this paper, we synthesize techniques from across various literatures to develop a pragmatic workflow that guides decision-making and identifies promising approaches for addressing common challenges in estimating dynamic models. We further illustrate the application of these estimation methods via two distinct example models, that showcase a range of different behavior modes. Our primary goal is to provide clear, concise guidelines for rigorously connecting System Dynamics (SD) models with quantitative data in order to determine reasonable parameter values or distributions, allowing for a match between simulated model variables and corresponding target data variables.