2011 Edward Tufte

I attended an excellent workshop by Edward Tufte, in Los Angeles last week. Below are some notes I took. He has published four books that have a wealth of information as well as amazing charts and data displays.


Data displays (charts)
Focus on causality; interaction of cause and effect
Lines = verbs, arrows = nouns, dotted lines = uncertain data
Causality is visually expressed on the (linking) lines. That’s where the action is.
Drives you to causal thinking (verbs not nouns)
Every design movement should carry content or data
Don’t order tables alphabetically, show in a logical order
Ask, what is your story?
Ask, why should they believe you?

When gathering data use ‘Whatever it takes’ principle
Have a spirit of inquiry when looking for data
Don’t cherry pick data
Chart chunk = replace large amounts of data or numbers and simplify
Segregate data by content, not by mode of production
Establish credibility: send links to data you find, source note, diversity of data
Show credibility by showing mastery of detail
Parenthetical: at least until better evidence or alternative explanation is found
You want an open mind to other possibilities, but not an empty head

Fundamental Principles of Analytical Design
(from Beautiful Evidence, Edward Tufte, pp. 122-139)
1. Comparisons (contrasts and differences)
2. Causality, Mechanism, Structure, Explanation
3. Multivariate Analysis (more than 1-2 factors)
4. Integration of Evidence (whatever it takes attitude)
5. Documentation (credibility)
6. Content Counts Most of All

Analytical presentation ultimately stands or fails depending on the quality, relevance and ingenuity of its content.

Strategy for non-fiction reports / presentations
Know your content, respect your audience
Engage audience using first person language (I, we); second (you) or third (they) person speaks down to the audience
Story + credibility
Analytical design comes from analytical thinking (mainly done to make comparisons)
Arrange information to help intellectual task
Cognitive tasks are turned into principles of evidence, presentation and design
Get rid of design; it’s an impediment between interface and viewer
Short / intense presentation
Increase information and throughput
Strip out jargon

How to make a presentation (cognitive style)
Print out Summary: problem, relevance, solution
Print out Technical Report, don’t show as a slide. High res data dump
Budget questions on another handout
Presentation points out the things that are important, follow by discussion
Meeting is about information consumption = good. We, the person, are not the center of attention.
Have really good content
Show up early (to the meeting you’re leading)
Finish early

Flow diagram
Intriguing texture of content is a sign of good design
Genuinely interactive, giving a cue to look further
High information (detail) shows access to design
People have a high capacity for information processing
Note: information travels the optic nerve at a rate of 20 megabytes / second

Enhance signal by reducing the amount of noise (other signals)
Don’t use legends or codes if you can avoid them (they impair learning)
For comparison, show things in adjacent space (instead of flipping)
Deep stacking is clunky and hard to follow
Use small multiples as an effective way to display information
Tables: recommend Gill Sans or Trebuchet

Interface design
Should be 90% content
Managing the flood of material and relevance of information
Fast scanner: yes, no, look again

Small word size graphics that display quantitative information
Show recent recording and last 500 (+/-) readings
Note: cognitive bias to put too much weight to most ‘recent’ information
Sparklines might not be technically accurate, but give the right answer to the question
Galileo’s first book used symbols (great change in cognitive thinking)

The publication ‘Nature’ has best data displays
The sports stats charts are also well done
Find a super graphic about your field
The publication ‘Library of Science’ is a beautiful journal
Napoleon’s March Chart was designed to be an anti-war poster
The Elements of Geometry, 16th Century, Euclid
Beautiful Evidence book (page 137, read this) Principles of Analytical Design