class: center, middle, inverse, title-slide # Data Visualization ## CORE Lab ### Department of Defense Analysis ### 2019-08-08 --- # Introduction * Purpose * Scope * Data visualization in R --- # Why visualize data? * Leverage capabilities and bandwidth of visual system to move information (often huge amounts) into the brain quickly. -- * Amplifies cognition in several ways (Card et al., 1999): * Increase cognitive resources * Reduce search * Enhance recognition of patterns and outliers * Support inferences of relationships * Can provide a medium with which to interact -- * We think visually. --- # Close Personal Network
--- # Close Personal Network
--- # Close Personal Network
--- # Close Personal Network
--- # Close Personal Network
--- # Common Mistakes * **Inappropriate use of visual cues/properties** * Too many dimensions * Lack of context * Attempting to substitute visual appeal for substance * Accidentially misleading audience * Quality of underlying data --- # Inappropriate use of visual properties .center[ ![](inapprop_vis_cues.png)] --- # Common Mistakes * Inappropriate use of visual cues/properties * **Too many dimensions** * Lack of context * Attempting to substitute visual appeal for substance * Accidentially misleading audience * Quality of underlying data --- # Too many dimensions .center[ ![](dimensions.jpg)] --- # Common Mistakes * Inappropriate use of visual cues/properties * Too many dimensions * **Lack of context** * Attempting to substitute visual appeal for substance * Accidentially misleading audience * Quality of underlying data --- # Lack of Context .center[ ![](context.png)] --- # Common Mistakes * Inappropriate use of visual cues/properties * Too many dimensions * Lack of context * **Attempting to substitute visual appeal for substance** * Accidentially misleading audience * Quality of underlying data --- # Visual appeal for substance .center[ ![](context.png)] --- # Common Mistakes * Inappropriate use of visual cues/properties * Too many dimensions * Lack of context * Attempting to substitute visual appeal for substance * **Accidentially misleading audience** * Quality of underlying data --- # Misleading audience <br> .center[ ![](mislead.png)] --- # Common Mistakes * Inappropriate use of visual cues/properties * Too many dimensions * Lack of context * Attempting to substitute visual appeal for substance * Accidentially misleading audience * **Quality of underlying data** --- # Quality of underlying data * Who? What? Where? When? Why? How? -- * Codebooks --- # General Principles and Guidelines * Answer questions and effectively communicate -- * Aesthetic, substantive, and perceptual considerations (Healy, 2019) * Gestalt laws of pattern recognition * Functionality over aesthetics * Choose right type of visualization * Avoid including extraneous data and dimensions (e.g., 3D) * Use colors appropriately * Label usage * Legends, titles, axes, scales, etc. --- # Example <br> .center[ ![](labels.png)] --- # Example .center[ ![](choropleth.png)] --- # Questions? <br> <br> <br> .center[ ![](corelogo5.png)] <br> <br> <br> <br> <br> <br> Dan Cunningham - dtcunnin@nps.edu