Narrative Visualizations: Depicting Accumulating Risks and Increasing Trust in Data

In contexts where people lack prior knowledge and risk awareness—such as the COVID-19 pandemic—even truthful visualizations of data can seem surprising. This can lead people to mistrust the veracity of the data and to discount it, leading to poor risk decisions. In this work, we illustrate how narrative visualizations can achieve a balance between the benefits of three common risk communication mediums (static visualizations, interactive simulations, and affect-laden anecdotes). We demonstrate empirically that a narrative visualization mitigates the reduced concern induced by a static visualization when communicating COVID-19 transmission risk (Experiment 1), and through mediation analysis we show that narrative visualizations increase trust via a different mechanism compared to anecdotes (Experiment 2). We argue that narrative visualizations deserve attention as a distinct class of visualizations that have the potential to be powerful tools for scientific communication (especially in contexts where data are surprising, and empiricism is important).