The popular notion that emotion and reason are incompatible is no longer defensible. Recent research in psychology and cognitive science has established emotion as a key element in numerous aspects of perception and cognition, including attention, memory, decision-making, risk perception, and creativity. This dissertation centers around the observation that emotion influences many aspects of perception and cognition that are crucial for effective visualization.First, I demonstrate that emotion influences accuracy in fundamental visualization tasks by combining a classic graphical perception experiment (from Cleveland and McGill) with emotion induction procedures from psychology (chapter 3). Next, I expand on the experiments in the first chapter to explore additional techniques for studying emotion and visualization, resulting in an experiment that shows that performance differences between primed individuals persist even as task difficulty increases (chapter 4). In a separate experiment, I show how certain emotional states (i.e. frustration and engagement) can be inferred from visualization interaction logs using machine learning (chapter 5). I then discuss a model for individual cognitive differences in visualization, which situates emotion into existing individual differences research in visualization (chapter 6). Finally, I propose an preliminary model for emotion in visualization (chapter 7).
