From a once-in-a-century pandemic to volatile swings in stock markets to turbulent political elections, uncertainty is all around us. Yet why do so many data analysts, data scientists and data journalists tend to ignore uncertainty? Or worse, can bad uncertainty representations lead to biased decisions? In this talk, we explore research from information visualization that endeavors to design better uncertainty representations to enable better decision-making and human-computer interaction. Building from theories in psychology, cognitive science, and behavioral economics, novel visualization design strategies for uncertainty like frequency framing with icon arrays and animated sampling in hypothetical outcome plots have shown great promise for improving uncertain predictions, mitigating cognitive biases, and may even be key in fostering trust in black box machine learning. This talk will first outline how visualization researchers have applied such new uncertainty visualizations to everyday situations of interpretating COVID vaccine efficacy to hurricane forecasting to investing for retirement. We’ll then focus on current research at UNCC’s Ribarsky Center for Visual Analytics to study how uncertainty visualizations can be used to enable data scientists to better communicate and measure uncertainty in their work.