Abstract for: Exploring Loop and Parameters Influence using Sensitivity Analysis with Explanatory Model Analysis

Sensitivity analysis is a valuable method for exploring system dynamics models, and an open challenge is to find new ways to help modellers to interpret the results, and discover influential model loops and parameters. Our approach uses a statistical learning approach known as explanatory model analysis in order to generate new insights from sensitivity data outputs. The process involves generating sensitivity data from a model, , building a regression model, and using analysis tools to generate model insights. Our results show that measurable results can be generated to show the impact of feedback loop policies and continuous valued parameters on an overall payoff value. For example, the most impactful feedback loops can be identified. We believe this method can generate useful results to contribute to the system dynamics modelling process. In terms of alignment, it could benefit two steps of an iterative approach to the modelling process, namely: step (4) testing and step (5) policy formulation and evaluation.