Actionable Uncertainty Visualization

Uncertainty has gotten a lot of attention from the visualization community; however, we are still very immature in understanding how to actually use the visual conveyance of uncertainty in quality decision making using science. Recent efforts have focused on various methods for showing uncertainty [1] or best practices for getting users at different technical levels to understand probabilities and integrate them into real-life decisions such as when to leave to catch a bus [2]. However, decisions using complex science to answer are much more difficult. Scientific investigations are typically formulated in a way to simply understand the science domain, and the results of those inquiries are not easily accessible to laypeople or policymakers [3]. The tools we are developing today in this space aim to convey uncertainty, variability, and sometimes error in the scientific data sets, but fail to encompass the needs of people using the science discoveries housed in that data for real decisions that make real impact. Without a rethinking of how we are transforming scientific data into visual forms that facilitate action, we are doing a disservice to the public that is funding our scientific work.