Communicating Uncertainty through Line Charts

Abstract

In recent years, the world has increasingly relied on mathematical models to predict the spread of COVID-19. This information is crucial for healthcare professionals in preparing to ensure that necessary care is available to patients when needed, for policymakers in making vital decisions and implementing policies to mitigate the spread of the disease, and for individuals in making important choices regarding their personal and professional lives. These prediction models can be highly intricate, considering numerous variables and scenarios. It is essential to acknowledge that the results of these models are inherently uncertain and never completely deterministic. Effectively communicating this uncertainty to users of the predictions is vital to ensure that informed decisions are made based on this information. Interpreting complex visualizations is known to be influenced by the visualization techniques as well as the individual differences of the users. Thus, this thesis conducts a comprehensive user study of how to effectively communicate uncertainty in time series prediction, such as COVID-19 prediction that are visualized as line charts. In addition to accurately interpreting uncertainty, this study also aims to assess whether it fulfils the informational needs of the users and whether the provided information is sufficient to motivate users to make informed decisions. If these needs are not adequately addressed, the thesis endeavours to understand why and explore ways in which they can be better met. This thesis is a part of ESID, which is an abbreviation for `Epidemiological Scenarios for Infectious Diseases´, a visual analytics application for epidemiological analysis, developed by the Institute for Software Technology at the German Aerospace Center (DLR).