Data Visualization Principles: Other Perceptual Channels CSC444 - - PowerPoint PPT Presentation

data visualization principles other perceptual channels
SMART_READER_LITE
LIVE PREVIEW

Data Visualization Principles: Other Perceptual Channels CSC444 - - PowerPoint PPT Presentation

Data Visualization Principles: Other Perceptual Channels CSC444 Acknowledgments for todays lecture: Tamara Munzner, Miriah Meyer, Colin Ware, Penny Rheingans History Time! Gustav Fechner, 18011887 Founder of psychophysics (What?)


slide-1
SLIDE 1

Data Visualization Principles: Other Perceptual Channels

CSC444

Acknowledgments for today’s lecture: Tamara Munzner, Miriah Meyer, Colin Ware, Penny Rheingans

slide-2
SLIDE 2

History Time!

Gustav Fechner, 1801–1887 Founder of psychophysics (What?)

slide-3
SLIDE 3

Psychophysics

  • Some stuff that happens in the “external

world” (outside your own body) causes stuff to happen “in your head”

  • Some of it is simple enough to study: that’s

psychophysics

  • “the scientific study of the relation between

stimulus and sensation"

slide-4
SLIDE 4

Stephens’s Power Law

Source: Stephens’s “Psychophysics”

slide-5
SLIDE 5

There exist stimuli

  • ther than colors
slide-6
SLIDE 6

From Stephens’s Psychophysics:

slide-7
SLIDE 7
slide-8
SLIDE 8
slide-9
SLIDE 9

So what is data visualization?

slide-10
SLIDE 10

The art and science of matching the “features” of a data set to the “features” of visual perception

slide-11
SLIDE 11

Why visualization?

slide-12
SLIDE 12
slide-13
SLIDE 13

Why visualization?

  • It has been studied more deeply
  • It appears to have more “bandwidth” than

alternatives (though not as much as you think it does)

  • It is richer
slide-14
SLIDE 14

(c) PlusMinus, GFDL

slide-15
SLIDE 15

THE STANDARD VISUAL CHANNELS

slide-16
SLIDE 16
slide-17
SLIDE 17

Cleveland/McGill perception papers

  • The beginning of visualization as an experimental

science

  • Required reading for ALL students!
slide-18
SLIDE 18

Cleveland/McGill perception papers

Better to worse:

slide-19
SLIDE 19

Pie Chart Bad, Scatterplot Good

slide-20
SLIDE 20

Integral vs. Separable Channels

  • Do humans perceive values “as a whole”, or “as

things that naturally split”?

slide-21
SLIDE 21

Integral vs. Separable Channels

(color, location) (color, motion) color x shape size x orientation x-size x y-size r-g x y-b

Colin Ware, 2004, p180 Separable Integral

slide-22
SLIDE 22

Bivariate Color Map (Bad)

Baraba and Finkner, via Tufte (VDQI)

slide-23
SLIDE 23

Bivariate Color Map (less bad)

http://www.csee.umbc.edu/~rheingan/636/color.pdf

slide-24
SLIDE 24

Trivariate (!) Color Map (terrible, terrible idea)

http://magazine.good.is/infographics/america-s-richest-counties-and-best-educated-counties#open

slide-25
SLIDE 25

The best bivariate colormap I know

http://www.nytimes.com/interactive/2014/11/04/upshot/senate-maps.html

slide-26
SLIDE 26

Bivariate Color Maps are Possible, but Hard

pay attention to the behavior of the variables you’re mapping from, and the behavior of the channels you’re mapping to.