Abstract: When are data visualizations persuasive, leading to changes in beliefs, attitudes, or behavior? The COVID-19 pandemic has starkly illustrated how visualizations can be powerful drivers of behavior change for some people, while having less impact for others when the data contradicts existing views or comes from untrusted sources. While there is broad recognition that perceptual, cognitive, and motivational processes affect how people interpret data visualizations, it is difficult to predict how the interplay of these factors determines whether visualizations prompt belief change and broader shifts in attitudes. In this talk I will argue that formal models of belief change provide a valuable quantitative framework for investigating how people make sense of and learn from data visualizations. I will describe recent work using Bayesian cognitive modeling to examine how people learn about statistical relationships (e.g., the correlation between mask-wearing and COVID infection rates) when interacting with data visualizations with different representations of statistical uncertainty. This modeling framework provides precise, testable predictions for how people should update their beliefs when interacting with data while accounting for perceptual biases, prior experience, and beliefs about the data-generating process. Finally, I will discuss potential insights from this approach and related work in cognitive science for the design and evaluation of persuasive data visualizations.
Bio: Doug Markant is a cognitive scientist who examines human learning, memory, and decision making through behavioral experiments and computational modeling. He obtained his Ph.D. in Psychology from New York University in 2014, after which he was a postdoctoral researcher at the Max Planck Institute for Human Development in Berlin in the Center for Adaptive Rationality. He is currently an Assistant Professor in the Dept of Psychological Science at UNC Charlotte, where he is a core faculty member of the Health Psychology Ph.D. program and an affiliate of the Ribarsky Center for Visual Analytics. He also serves as the director of the interdisciplinary program in Cognitive Science.