1. Select the right graph.
| What to plot? | Common graphs to use |
|---|---|
| Distribution of a continuous variable | * Histogram or kernel density curve to visualize the distribution * Q-Q or P-P plot to check normality |
| Distribution of a categorical variable | * Bar graph * Pie chart |
| Relation between a continuous variable and a categorical variable | * Box plot * Bar graph * Line graph (when the categorical variable has order) |
| Relation between two continuous variables | * Scatter plot * Line graph (usually put the time variable on the x-axis) |
| Relation among more than two variables | * Be creative! |
2. Tells a good story with your data.
- Examples: More Americans are tying the knot
3. Avoiding too much information in one graph. Simplify less important information.
- Examples: Ruining the cake with too much icing
4. Highlight what is important.
- Examples: Doing the line charts right
5. Do not burden your audience with visual math.
- Examples: Visual math goes wrong
6. Select appropriate colors.
- Examples: Being clever with color
7. Being creative with your legends and labels.
- Examples: My penchant for horizontal bar graphs
References and Resources