Collaborative Research: HCC: Medium: Responsive Uncertainty Visualizations for Human-Centered Decision-Making

Visualizations of data, such as charts and graphs, are increasingly used to communicate and analyze data in business, health, journalism, and many other domains. When presenting data, one important issue is uncertainty: data can be noisy, incomplete, or biased, while statistical analyses based on data often return a range of possibilities. Failing to include information about uncertainty in a visualization can mislead viewers; however, adding uncertainty information can make visualizations more complex, which could also cause confusion. The key idea of this project is that visualizations should be designed to be responsive to the needs, abilities, and resources of viewers, displaying an appropriate amount and complexity of uncertainty information. For instance, someone with experience in data analysis and facing a long-term strategic decision might want a complete picture of the range and likelihood of outcomes and could afford to spend time poring over a complex set of contingencies and predictive intervals. Someone with less experience and making a short-term tactical decision might get the most utility from a simpler representation that shows the estimated average case, along with the lower and higher boundaries of the estimate. The project team will develop methods to model people's abilities and needs, along with a library of ways to represent uncertainty, that can be combined to make more valuable and widely accessible visualizations that improve people's decision-making ability.The goal of the project is to explore the design space of responsive uncertainty visualization. To do this, the research team will first examine the relevant resources needed when considering visualization design, from practical concerns like screen display size and internet bandwidth to higher-level concepts like visualization literacies, decision-making constraints, and rhetorical goals. This will be done in conjunction with stakeholders from diverse backgrounds who work in high-impact decision scenarios like crisis informatics. The team will then explore how designs for uncertainty visualizations can be adapted to the constraints imposed by these various resources, creating a "design dictionary" of responsive approaches, ready for use for any designers working to communicate uncertainty information to diverse audiences.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.