Principles of a Good Visualization

Edward Tufte is an American Statistician who published a book named Visual Display of Quantitative Information. In this book Tufte mentions some principles which are considered gold standard for good visualizations by many. In this report we are going to discuss these principles.

Tufte’s principle of Graphical Integrity

Following are the six Tufte’s principles of Graphical Integrity:

1. The representation of numbers, as physically measured on the surface of the graph itself, should be directly proportional to numerical quantities measured.
2. Clear, detailed and thorough labelling should be used to defeat graphical distortion and ambiguity. write out explanations of the data on the graph itself. Label important events in the data.
3. Show data variations not design variations
4. In time-series displays of money, deflated and standardized units of monetary measurement are nearly always better than nominal units.
5. The number of information-carrying (variable) dimensions depicted should not exceed the number of dimensions in the data.
6. Graphics must not quote data out of context

Tufte’s Principle of Design

• Above all show data.

Here comes the concept of Lie factor as described by Tufte’s.

Lie factor can be summarized as ratio of the size of the graphic in the visualization to the size of the data itself. A ratio of 1, shows the visualization is able to represent the real nature of data. Any deviation from 1, either undermines or exaggerates the effect. Another term described by Tufte’s is Chart Junk. Chart Junk is the extra effects on the visualizations that do not add any new information to the graphics. They just consume space and create more clutter. Grid lines usually come in chart junk. Grids should usually be muted or suppressed in the final visualization because grids are meant for initial plotting and not for final version.

• Maximize data ink ratio

Data ink ratio can be described as the ratio of the non-erasable core ink in a graphic visualization to the total ink used in the visualization. If the data ink is removed, then valuable information is lost. In order to remove chart junk, we should aim for higher data ink ratio.

• Erase non data ink

If there is loss of information when erasing some graphics then that is called as core ink or non-redundant ink. Remove non data ink to improve the data ink ratio and get better visualization.

• Erase redundant data ink

If there is repetition of information or labels than that particular data ink needs to be removed as they are redundant and don’t add to any additional information for the user.

• Revise and edit

Prior to publishing, we should carry out the above steps again to revise and make the visualization better for viewers.

References:

1. The Visual Display of Quantitative Information by Edward Tufte
2. Figure 1 taken from https://analythical.com/blog/covid19-in-charts
3. Figure 2 taken from https://www.businessinsider.com/the-27-worst-charts-of-all-time-2013-6?IR=T
4. Figure 3 from Principles of Data Visualization by James Bernhard
5. Figure 4 taken from https://www.whattheviz.com/post/show-the-data-understanding-data-ink-ratio

1 Response

1. Anupama says: