Publisher: Graphics Press
Copyright: 1990
ISBN: 0961392118
Envisioning Information is very similar to another of Tuft's books I read, The Visual Display of Quantitative Information. Envisioning Information doesn't focus as much on the display of data per se but rather broadens the scope to include all types of information. As you would expect with a book on such a topic, this one is a visually pleasing experience. The included graphs, maps and pictures along with the anecdotal evidence for Tufte's conclusions made this an enjoyable read.
Flatland and Floods
Tufte uses the term flatland to describe the 2 dimensional world onto which we transcribe data that lives in our 3 dimension plus time world. The amount of data we try to transcribe often determines the method we use and is in turn often determined by how long we have been collecting a particular type of data.
For instance when Galileo first began charting sunspots, he had only a few samples from his new telescopic arrangement. He used a spot diagram to first show what he was seeing. Then he became interested in the changes over time so he used a small multiple chart... the same diagram repeated on the same page with only the differences showing.
Over time however scientists built up hundreds of thousands of observations and in 1904 a man by the name of E. W. Maunder drew a chart of sunspots that became known as the butterfly chart. His chart covered hundreds of observations over several decades and showed the cyclical nature of sunspots on the sun's surface.
Now we have charts that show this effect over a hundred years times and on this chart, Tufte superimposes some macro level data regarding not just the number of spots recorded but their relation to the equator of the sun and what percentage of the sun's surface was covered by spots.
When data goes from a trickle to a flood, there is a series of strategies used to aggregate and display the data. Tufte delineates this progression as such:
- Spot diagrams
- Small multiple diagrams
- Dimensionality and data compressions
- Micro / macro displays combining patterns & details, averages & variations.
The last bit was emphasized considerably by Tufte. There are two fundamental summary measures in statistical data: averages and the variation about the average. Charts need to show these two things in order to be effective.
Notes & Quotes
"If the numbers are boring then you've got the wrong numbers."
"To clarify, add detail."
"Clutter and confusion are failures of design, not attributes of information."
"Information consists of differences that make a difference."
Re: Grids. Try to do away with the lines of a grid if at all possible. Allow the data to draw the lines.
Re: Use of color. There are 4 major uses of color:
- Label
- Measure
- Represent (imitate)
- Decorate
Conclusion
I love reading books like this. Short and yet packed full of useful information. If you are involved in design in any way, whether it is full fledged graphic design or just making a newsletter for your local organization, you should read this book and frankly anything you can get your hands on that Tufte wrote.