For an overview of pitfalls to avoid in ethical, visual representation of data, see the pages on Pitfalls in dataviz: scales, Pitfalls in dataviz: colours and Pitfalls in dataviz: chart types. These pitfalls are listed here again for reference. So, ethical data visualisation also means
- not breaking scales
- choosing an appropriate number to start your y axis in line scales
- choosing an appropriate width to height ratio
- not distorting proportions
- preserving the same scale for comparisons
- avoiding rainbow colour scales and using perceptually uniform colour scales
- using appropriate chart types for your data and message
- avoiding double y axis charts
- choosing an appropriate line interpolation for line charts
- being careful with stacked area and stacked bar charts
- being careful no to suggest causality when there is only correlation
These pitfalls should be avoided to produce undistorted, meaningful and truthful visualisations to communicate data.
When any chance in misinterpretation arises, chart authors should guide the reader in a way that visualisations are clear, and can only be interpreted in the correctly. This can be done by making good use of labels, axis titles, legends and annotations.