2.19 Visualization Best Practices

A well-known tenet of data modeling is the parsimony principle – never use more than the minimum required number of inputs when building the model.  Do not increase complexity for its own sake.  We recommend applying the same general concept to visualizations.  

With any expository visualization that you build, your primary goal is to communicate to your audience.  

Never assume that the meaning of a variable in your dataset is obvious to your audience.  Use axis labels and titles to improve clarity.  

When generating visualizations, it is essential to pay attention to variables’ types.  

Finally, always remember that effective data visualizations may generate more questions than answers – and that’s completely okay!