Transparent Risks Revisited: Evidence for a Dark-is-More Bias in Risk Perception

Prior research has shown that different representations of uncertainty in data visualizations can lead to more (or less) riskaverse decision making. It is crucial for researchers to develop a better scientific understanding of these effects so that visualizations such as hazard maps can be designed to support viewers in reasoning about risk and probability. This paper presents a follow-up to a prior study that showed that participants underestimated the risk from a wildfire when transparency was used to represent different risk levels. In the present study, we test the hypothesis that the participants' decisions about risk are influenced by the dark-is-more bias. Across three experiments using the same wildfire evacuation task, we found that participants were consistently more likely to evacuate when the probability bands representing the fire risk were darker.