Data Presentation

The baseline and scale of your charts are important factors to consider in order to accurately represent your data. With the possible exception in analysis of economic and finance data, most charts should have a natural baseline of zero. However, using a baseline that is logical, meaningful, and that does not misrepresent the data.

Here are a few examples of how non-zero baselines have been used to alter the perception of data:

This first chart is from a Fox News broadcast.

The first chart was provided from a 2012 Forbes article.

This second chart uses the same data, with a baseline of zero. Notice that the difference in tax rates appears much less drastic. 

The second, corrected chart is from flowingdata.com.

This is another example from a 2017 Vox story

It is not enough to choose an appropriate visualization type, color scheme, and scale. It is also important to consider how you will organize your data within a visualization.

Do you have wide ranges? Perhaps using multiple charts will better illustrate the variations in different scales.

Do you have a large quantity of data? If so, you may try changing chart types, splitting up data points, or simplifying colors so that your chart doesn't become too overwhelming.

Are you comparing multiple sets of variables? It may be helpful to try several chart types to make sure your data is easy to read while conveying the information you desire. Don't be afraid to break visualization conventions for chart types if your data is better visualized using a non-standard method.