Rauser Hightlights

Assembly - The grouping of detected graphical elements

• reification: You experience more spatial information than is present in an image.
• emergence: You recognize elements by a complete appearance. Independent of the dots. (28:00)
• pragnanz (pithiness): We strongly prefer to interpret stimuli as regular, simple, and orderly (28:55)
  • Law of Closer: We combine points into a picture (29:00)
  • Law of continuity: We group together objects that follow an established direction (29:12)
  • Law of similarity: We tend to see elements that are physically similar as part of the same object (30:27) shape, color, degree of curvature
  • Law of proximity: We tend to group objects that are near each other into a group (32:05)
• Weber’s Law: The “Just Noticeable Difference” is proportional to the size of the initial stimuli (35:24)
  • 10 vs 20 is easy to discriminate, but 100 vs 110 is much more difficult
  • White bars and blue bars (36:32)

My notes from the Presentation

1. pre-attentive processing: Subconscious understanding of graphs.
2. Only plot time-series if you have sequential data. (3:16)
3. Bill Cleveland book (4:00)
   • Detection
   • Assembly
   • Estimation
4. Table look up is a very bad thing (9:47)
5. Alphabetical sorting in a plot is almost always bad (10:27)
6. Stacked bar charts are almost always poor.
7. Just use two charts (I have this problem) (18:51)
8. Pie charts are always a mistake (19:36)
9. Comparison is trivial on a common scale (21:44)
   • “Two table look ups and a slot.”
10. Standardize: Subtract the mean and divide by the standard deviation
11. The dashboard metaphor is fundamentally flawed (23:45)
    • This is the argument against software like tableau and maybe Excel’s instant plot function
12. Use two channels to encode information redundantly. This is often a good choice.
13. Dodged bar charts are often a bad idea (32:49)
14. Excel’s defaults are pretty bad (34:35)
15. Weber’s law is why grid lines are useful (36:48)
16. Erase non-data ink, with reason – Tufte (37:23)
17. You are best at detecting slope near 45 degrees – banking to 45 (38:10)
18. Should I include zero in my scale – it depends (39:02)
    • bar charts are comparing lengths. The axis has to start at zero.
19. Above all else show the variation in data – Rauser (40:32)

Piecharts are the information visualization equivalent of a roofing hammer to the frontal lobe. They have no place in the world of grownups, and occupy the same semiotic space as short pants, a runny nose, and chocolate smeared on one’s face. The are as professional as a pair of —less chaps.

Coda Hale