Data Visualization for Speakers

The wrong chart type can obscure the very insight you are trying to communicate. A pie chart with twelve slices is unreadable; a line chart showing trends with only two data points is misleading. The choice depends on what you want to show: comparisons between categories call for bar charts, trends over time need line charts, part-to-whole relationships suit stacked bars or treemaps, and correlations between variables use scatter plots. Before opening any software, ask yourself: 'What is the one insight I want the audience to take from this data?' That answer determines the chart type, the highlighted element, and the annotation you add.