It's been a while Data visualization handbook by Juuso Koponen and Jonatan Hildén has been waiting on my book shelf. I started to go through it about a week ago and here are a few quotes that stayed with me.
Notes before we start:
- The page numbers I use are from the first edition (2020).
- Some quotes might only be understandable within context.
- I will add to this post until I have finished the book.
Text v. data visualization:
When should data be visualized, and when is text enough? In short, if something can be expressed just as clearly or even more clearly in words, visualization is unnecessary—sometimes even couterproductive. (p. 29)
[There is] one rule that an information designer should always follow: choose the clearest presentation method available. Disregard any other rules and guidelines we present in this book, when they are in conflict with this golden rule of information design. (p. 43)
At its best, a visualization gives an overview of the topic quickly, but also rewards the reader who spends time exploring it in depth. (p. 43)
General principles of visualization design
Focal points guide the eyes:
With the use of focal points1, the designer can introduce linear features into an open narrative structure. If it is important that specific elements be read before others, these elements should be highlighted visually so that the reader's gaze is drawn to them first. Highlighting key details is also justified by the fact that it makes it easy for the reader to understand which of the features in the graphic are the most essential. (p. 90)
Integrate text and graphics:
In an eye-tracking study conducted at Lund University, the subjects were presented with newspaper spreads that had identical content, with the text and graphics laid out in different ways. [...]
The study also looked at an alternative, integrated layout, in which the graphic elements were placed withing the text column, so that the teext and graphics that are related in content were located close to each other. When a page was laid out like this, readers read both the body text and the related graphics. This solution places less cognitive strain on the readers, because they do not have to form mnemonic links between the text and the separate graphics. (p. 90)
Nominal v. ordinal scales:
A nominal scale divides items into classes, the only relevant characteristic of which is that they are separate. They have no particular order or distance on the scale. Examples of such classes include nationality, profession, membership in a certain group, and so on. [...] Values on a nominal scale cannot be mathematically compared. [...]
An ordinal scale differs from the nominal scale in that the classes have a rank order: the first to tenth income decile, the Likert scale (strongly agree, agree, etc.), a ranking in comparison or competition, a military rank, and so on. It is possible to determine the median value for items on an ordinal scale, and to calculate the difference between two rankings, but because the scale does not say anything about the magnitude of differences between classes, most other calculations are not meaningful. (p. 94)