The following blog is designed to summarize Chapter 1 of Edward R Tufte’s The Visual Display of Quantitative Information on Graphical Excellence. The chapter is then condensed into 3 key principles and then finally I give reflections on the chapter as a whole.
Graphics should:
Show the data, make the viewer think about the substance rather than the design, production or method, avoid distortion, summarize large and complex data sets, encourage comparison, reveal different levels of detail, serve a clear purpose.
Maps are effective at illustrating large amounts of regional data, encouraging comparison and are adaptable to different levels of detail. A drawback is that in some cases a map is limited by its own composition – maps are composed of regions of a set size with observers eyes potentially drawn to the larger regions even if they are less significant to the story of the data.
Time series charts are good at exploring data with a large time range and can reveal variability in trends over time or comparison between groups. A limitation might be that time is often not a satisfactory explanatory variable, often times we might be able to identify a change with time series but need other graphics to help explain it. A before-after time series offers potential for more explanatory observations.
The Principles of graphical excellence according to Tufte:
A well designed presentation of data that balances substance, statistics and design.
Complex ideas communicated with clarity, precision and accuracy.
Giving the viewer the greatest number of insights and ideas in the shortest time with only what is necessary with regard to information and space.
Reflections:
I enjoyed reading the first part of this book as it referred back to some of the themes in my first blog about data stories. For example, revealing different levels of detail is closely linked with a drill down story. I also liked that there was praise for design but not without the major qualifier that it cannot be at the expense of substance and Chapter 2 further explores the problem of integrity when design might begin to distort substance. Ultimately it returns to the core principle of clarity in any visualization; design should enhance clarity not compromise it.