There is no one-size-fits-all datavis, but most books are written from a very generic or a graphic design perspective, Stephen Few’s books being one of the exceptions

[update: should have also mentioned Naomi Robbins’s book]. So, I decided to write an unsexy entry level data visualization book for the illiterate and artistically challenged spreadsheet user (I also wanted to know if I could come out with a reasonably consistent vision). It is oh-so-late-but-almost-done-now, so I decided to share with you its structure and hope you’re kind enough to send some feedback.

The constraints

From the beginning I defined some constraints:

  • The book is all about charts, no maps, no networks.
  • Every concept must be illustrated by a concrete example.
  • Real-world data.
  • All charts presented in the book must be made in Excel and unpolished by illustrator or something.
  • It’s a datavis book: it uses Excel but will not teach how to make charts in Excel.
  • It’s an ebook (for the moment) so no worries about printing costs.

Table of contents

Probably there are no surprises here, but the sequence may be a little unexpected. The whole book is structured around a few ideas discussed in chapters 1 and 2, one of them illustrated by the image above (is it a landscape model a simple area chart? Does it matter?).

  1. The building blocks. Where I discuss aliens, Magritte’s pipe and this idea of using geometric primitives and visual variables to design visual landscapes of abstract concepts. An example shows how a relatively common visualization (comparison of two pie charts) can perform so poorly when compared to a much simpler but effective slope chart.
  2. Human perception. Basic physiology of the eye, gestalt laws, short memory and pre-attentive processes. Cleveland, Weber, Stevens, and what to do minimize perceptual issues.
  3. Beyond perception. “Social pragnanz”, culture, rules and where to break them, graphical literacy, corporate culture, specialized contexts.
  4. Data preparation. Most Excel users are not aware of how important it is to have a well structured table. This chapter discusses this and basic ETL.
  5. What is information visualization. Definition, of datavis. Point, pattern and outlier detection tasks. From points to shapes and patterns. Skills.
  6. From exploration to explanation. The process from asking a question to communicating the results.
  7. How to choose a chart. Based on my classification of chart types.
  8. Comparison. Bars, slopes, points.
  9. Composition. Pies and what to do about them.
  10. Distribution. Scatterplots, histograms.
  11. Evolution. Lines, horizon charts and connected scatterplots.
  12. Relationship. Scatterplots.
  13. Profiling. Scatterplot matrixes, small multiples, horizon charts, trellis.
  14. Charts and graphic landscape. Dashboards. Multiple charts for multiple perspectives.
  15. Aesthetics in information visualization. After the data is ready, everything is design, but you don’t have to be designer to make effective charts. Emotion, reason of both?
  16. Color: how to avoid a catastrophe. What is color. How to use stimulus intensity to minimize color clashes and prioritize your message. Color and visual tasks. Here is the color wheel if you need it, but color harmony is just nice-to-have.
  17. How to format a chart. Scales, legends, titles, 3D. Some lies.
  18. Visualization in an organizational context. Exploration and communication cycles, effectiveness vs. impression management.
  19. Microsoft Excel as a visualization tool, and beyond. How to avoid Excel defaults, beyond the library, when and where to go next.

So, is something missing? Please share it in the comments below.