creating more effective charts · 2019. 7. 31. · poultry production is rising in australia...
TRANSCRIPT
CREATING MORE EFFECTIVE CHARTSEcho Swinford
Presentation consultant@echosvoice | echosvoice.com
Sensory Long-termWorking (ST)
Source: Cliff Atkinson, Beyond Bullet Points; Richard E. Mayer, Ed., The Cambridge Handbook of Multimedia Learning
Source: Cliff Atkinson, Beyond Bullet Points; Richard E. Mayer, Ed., The Cambridge Handbook of Multimedia Learning
Sensory Long-term
Gestalt principles
PROXIMITYObjects close together
are perceived as related.
FOCAL POINTSObjects with a point of interest, emphasis
or difference will capture attention
FIGURE-BACKGROUNDObjects are perceived as figures
(in focus) or background.
CONTINUITYObjects moving in the same direction
are perceived as related.
SIMPLICITYPeople will perceive and interpret ambiguous or complex images as the simplest form(s) possible
CONNECTEDENESSConnected objects
are perceived as related.
CLOSUREOpen structures are perceived as closed.
SIMILARITYObjects similar in nature are perceived as related.
https://www.smashingmagazine.com/2014/03/design-principles-visual-perception-and-the-principles-of-gestalt/
Pre-attentive processing
POSITION LENGTH ANGLE DIRECTION
SHAPES SATURATION COLOR
SIZE
ENCLOSURE LINE WEIGHT
https://policyviz.com/product/core-principles-of-data-visualization-cheatsheet/
Year Sheep & goat Beef & Buffalo Pig Poultry
1961 583,775 642,924 110,066 49,388
1962 596,526 803,950 122,328 49,740
1963 603,234 928,606 116,201 51,520
1964 595,377 1,001,315 113,190 60,384
1965 594,473 1,026,287 122,061 72,946
1966 608,689 946,333 135,105 69,316
1967 596,671 878,683 142,274 88,845
1968 664,747 904,285 149,438 102,929
1969 679,987 934,766 162,610 109,297
1970 755,175 1,009,954 174,684 124,384
1971 825,519 1,047,200 181,852 151,992
1972 956,703 1,164,393 194,118 163,741
1973 713,770 1,437,944 236,271 161,248
1974 467,847 1,321,817 211,048 193,982
1975 527,523 1,546,965 175,126 189,760
1976 588,095 1,840,415 173,989 204,552
1977 549,636 1,987,800 185,041 218,626
1978 513,995 2,183,800 199,097 246,487
1979 491,775 2,018,000 198,512 271,938
1980 548,875 1,564,400 219,580 313,457
1981 578,600 1,467,200 233,161 304,228
Glanceable vs referenceable
Source: Pew Research Center. Pew Research Center bears no responsibility for interpretations presented or conclusions reached based on analysis of the data
Poultry production is rising in Australia
Source: Pew Research Center. Pew Research Center bears no responsibility for interpretations presented or conclusions reached based on analysis of the data
Sheep & goat746,852
Beef & Buffalo2,586,317
Pig360,893
Poultry1,098,482
0K
500K
1,000K
1,500K
2,000K
2,500K
3,000K
1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 2009 2013
ton
nes
Glanceable vs referenceable
Source: Pew Research Center. Pew Research Center bears no responsibility for interpretations presented or conclusions reached based on analysis of the data
Sheep & goat746,852
Beef & Buffalo2,586,317
Pig360,893
Poultry1,098,482
0K
500K
1,000K
1,500K
2,000K
2,500K
3,000K
1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 2009 2013
ton
nes Year Sheep & goat Beef & Buffalo Pig Poultry
1961 583,775 642,924 110,066 49,388
1962 596,526 803,950 122,328 49,740
1963 603,234 928,606 116,201 51,520
1964 595,377 1,001,315 113,190 60,384
1965 594,473 1,026,287 122,061 72,946
1966 608,689 946,333 135,105 69,316
1967 596,671 878,683 142,274 88,845
1968 664,747 904,285 149,438 102,929
1969 679,987 934,766 162,610 109,297
1970 755,175 1,009,954 174,684 124,384
1971 825,519 1,047,200 181,852 151,992
1972 956,703 1,164,393 194,118 163,741
1973 713,770 1,437,944 236,271 161,248
1974 467,847 1,321,817 211,048 193,982
1975 527,523 1,546,965 175,126 189,760
1976 588,095 1,840,415 173,989 204,552
1977 549,636 1,987,800 185,041 218,626
1978 513,995 2,183,800 199,097 246,487
1979 491,775 2,018,000 198,512 271,938
1980 548,875 1,564,400 219,580 313,457
1981 578,600 1,467,200 233,161 304,228
Emphasis
Emphasis
Emphasis in charts
SATURATION COLORSIZEPOSITION/PROXIMITY
SHAPE
Background and foreground
5
100
15 20 25 30 35 40 45 5010
200
300
400
500
600
700
5
100
15 20 25 30 35 40 45 5010
200
300
400
500
600
700
Background
Foreground
When everything is bold, nothing is bold
0%
10%
20%
30%
40%
50%
60%
70%
Percent positive for salmonella
Ground chicken Young chicken Target Target
When everything is bold, nothing is bold
Ground chicken
Young chicken
Target
Target0%
10%
20%
30%
40%
50%
60%
70%
CY06Q1 CY07Q1 CY08Q1 CY09Q1 CY10Q1 CY11Q1 CY12Q1 CY13Q1
Color choices make a difference
Stephen Few, Perceptual Edge 2008
0
2
4
6
8
10
12
France UnitedKingdom
Denmark Australia Egypt
0
2
4
6
8
10
12
France UnitedKingdom
Denmark Australia Egypt
Color choices make a difference
Stephen Few, Perceptual Edge 2008
0
2
4
6
8
10
12
France UnitedKingdom
Denmark Australia Egypt
0
2
4
6
8
10
12
France UnitedKingdom
Denmark Australia Egypt
0
2
4
6
8
10
12
France UnitedKingdom
Denmark Australia Egypt
Stephen Few, Perceptual Edge 2008
Using too much color undermines its power
Chart junk
Any element of a visualization that is purely decorative, redundant, or unnecessary for understanding the data.
Extreme examples of chart junk
https://www.youtube.com/watch?v=_ZaGNPiZXhE
Not-so-extreme examples of chart junk
https://www.youtube.com/watch?v=_ZaGNPiZXhE
$0
$500,000
$1,000,000
$1,500,000
$2,000,000
$2,500,000
$3,000,000
$3,500,000
$4,000,000
$4,500,000
$5,000,000
January February March April May June July
Ave
rage
Inco
me
Months
Average Income Per Month
January
February
March
April
May
June
July
Remove chart junk
$0K
$1M
$2M
$3M
$4M
$5M
January February March April May June July
Average income per month
https://www.youtube.com/watch?v=_ZaGNPiZXhE
Removechart junk
$4.3M
$2.5M
$3.2M
$4.5M
$2.7M
$1.7M
$3.8M
January February March April May June July
Average income per month
https://www.youtube.com/watch?v=_ZaGNPiZXhE
$0K
$1M
$2M
$3M
$4M
$5M
January February March April May June July
Average income per month
24
$0
$500,000
$1,000,000
$1,500,000
$2,000,000
$2,500,000
$3,000,000
$3,500,000
$4,000,000
$4,500,000
$5,000,000
January February March April May June July
Ave
rage
Inco
me
Months
Average Income Per Month
January
February
March
April
May
June
July
Labeling and titling
Legends vs direct labeling
Widgets
Spinners
Geegaws
Thingies
Sprockets
Legends vs direct labeling
54%
21%
9%
8%
8%
Widgets Spinners Geegaws Thingies Sprockets
Widgets54%
Spinners21%
Geegaws9%
Thingies8%
Sprockets8%
Legends vs direct labeling
Widgets54%
Spinners21%
Geegaws9%
Thingies8%
Sprockets8%
Widgets54%
Spinners21%
Geegaws9%
Thingies8%
Sprockets8%
Legends vs direct labeling
0
1
2
3
4
5
6
JAN FEB MAR APR
Gizmo Sales (millions)
Widgets Sprockets Gizmos
0
1
2
3
4
5
6
JAN FEB MAR APR
Gizmo Sales (millions)
Widgets
Sprockets
Gizmos
Legends vs direct labeling
0
1
2
3
4
5
6
JAN FEB MAR APR
Gizmo Sales (millions)
Widgets Sprockets Gizmos
Widgets 4.5M
Sprockets 2.8M
Gizmos 5.0M
0
1
2
3
4
5
6
JAN FEB MAR APR
Gizmo Sales
Use McKinsey-style titles
Widgets 4.5M
Sprockets 2.8M
Gizmos 5.0M
0
1
2
3
4
5
6
JAN FEB MAR APR
Gizmo sales are rising
Widgets 4.5M
Sprockets 2.8M
Gizmos 5.0M
0
1
2
3
4
5
6
JAN FEB MAR APR
Gizmo Sales
Annual sales
Q1 Q2 Q3 Q4
Sorting adds meaning and clarity
Stephen Few, Perceptual Edge 2008
Alphabetical orderTOTALLANDAREA
Alabama
Alaska
Arizona
Arkansas
CaliforniaColorado
ConnecticutDelaware
FloridaGeorgia
Hawaii
Idaho
Illinois
IndianaIowa
Kansas
Kentucky
Louisiana
Maine
MarylandMassachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
NebraskaNevada
NewHampshire
NewJersey
NewMexicoNewYork
NorthCarolinaNorthDakota
OhioOklahoma
Oregon
Pennsylvania
RhodeIsland
SouthCarolinaSouthDakota
Tennessee
Texas
Utah
Vermont
VirginiaWashington
WestVirginia
Wisconsin
Wyoming
POPULATION
What kind of chart?
Bars and lines tell time differently
0
500
1000
1500
2000
2500
3000
3500
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2007 Dropouts
Actual Forecast
0
500
1000
1500
2000
2500
3000
3500
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2007 Dropouts
Actual Forecast
Stephen Few, Perceptual Edge 2008
Bars and lines tell time differently
0
500
1000
1500
2000
2500
3000
3500
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2007 Dropouts
Actual Forecast
Stephen Few, Perceptual Edge 2008 37
Bars and lines tell time differently
0
500
1000
1500
2000
2500
3000
3500
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2007 Dropouts
Actual Forecast
Stephen Few, Perceptual Edge 2008 38
Just because it equals 100% doesn’t mean it has to be a pie
Freshwater fish142
Cats88.3
Dogs74.8
Small animals24.3
Birds16
Horses13.8
Reptiles13.4
Saltwater fish9.6
https://www.stevedalepetworld.com/blog/world-pet-population-data-mixed-bag/
Just because it equals 100% doesn’t mean it has to be a pie
142
88.3
74.8
24.3
16 13.8 13.49.6
Freshwaterfish
Cats Dogs Small animals Birds Horses Reptiles Saltwater fish
https://www.stevedalepetworld.com/blog/world-pet-population-data-mixed-bag/
Just because it equals 100% doesn’t mean it has to be a pie
142 Freshwater fish 88.3 Cats 74.8 Dogs 24.3 Small animals
16.0 Birds
13.8 Horses
13.4Reptiles
9.6 Saltwater fish
https://www.stevedalepetworld.com/blog/world-pet-population-data-mixed-bag/
Just because it equals 100% doesn’t mean it has to be a pie
Freshwater fish 142
Cats 88.3
Dogs 74.8Small animals 24.3
Birds 16.0Horses 13.8Reptiles 13.4Saltwater fish 9.6
https://www.stevedalepetworld.com/blog/world-pet-population-data-mixed-bag/
Just because it equals 100% doesn’t mean it has to be a pie
https://www.stevedalepetworld.com/blog/world-pet-population-data-mixed-bag/
Ann K. Emery’s rules for using pie charts
1. Are well-formatted
2. Display nominal variables
3. Add to 100%
4. Contain positive numbers
5. Display a single point in time
6. Only have two or three slices
7. Are displayed individually
44https://depictdatastudio.com/when-pie-charts-are-okay-seriously-guidelines-for-using-pie-and-donut-charts/
No exploding pies
https://flowingdata.com/2012/04/25/world-happiness-report-makes-statisticians-unhappy/
Freshwater fish
Cats
Dogs
Small animals
Birds
Horses
Reptiles
Saltwater fish
Must add to 100%
https://flowingdata.com/2012/10/11/not-enough-donut/
Strongly Agree20%
Agree40%
Disagree30%
Strongly Disagree
10%
No sequential data
Ordinal / sequential data Ranges and other groupings
https://depictdatastudio.com/when-pie-charts-are-okay-seriously-guidelines-for-using-pie-and-donut-charts/
20% 40% 30% 10%
Strongly agree Agree Disagree Strongly
Disagree
0-920%
10-1940%
20-2930%
30-3910%
20%
40%
30%
10%
0-9 10-19 20-29 30-39
No multiple pies
https://depictdatastudio.com/when-pie-charts-are-okay-seriously-guidelines-for-using-pie-and-donut-charts/
Women20%
Men80%
Company A
Women40%
Men60%
Company B
Women60%
Men40%
Company C
Women80%
Men20%
Company D
20% 80%Company A
40% 60%Company B
60% 40%Company C
80% 20%Company D
Women Men
Slope charts are often a good alternative to pies
https://www.bloomberg.com/news/articles/2018-12-21/hong-kong-s-lost-decade-as-shown-by-these-12-global-surveys
Slopes vs pies
https://www.economist.com/graphic-detail/2012/07/03/bank-profits-head-east
peltiertech.com/slope-graphs-in-excel/
Slopes vs pies
https://www.economist.com/graphic-detail/2012/07/03/bank-profits-head-eastpeltiertech.com/slope-graphs-in-excel/
Pre-Tax profits of the 1000 largest banks (% of total)
Try a bullet chart instead of a column
0
500
1000
1500
2000
2500
3000
3500
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2007 Dropouts
Actual Forecast
0
500
1000
1500
2000
2500
3000
3500
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2007 Dropouts
Forecast Actual
Keys to creating effective charts
Leverage gestalt & pre-attentive attributes
Strip chart junk Simplify colors
Use direct labelingInclude the takeaway
in the titleChoose the right chart