Graphical Excellence

"Excellence in statistical graphics consists of complex ideas communicated with clarity, precision, and efficacy."

According to Edward Tufte, graphical displays should

show the data
induce the viewer to think about the presence, not the presentation
avoid distorting the data
present many numbers in a small space
make large data sets coherent
encourage the eye to compare different pieces of data
reveal the data in multiple levels of detail
serve a reasonably clear purpose
be closely integrated with the statistical and verbal descriptions of the dataset

Graphical excellence

Graphic excellence is "the efficient communication of complex quantitative data".

Since the 17th century, maps and statistics have been combined to understand what is happening in relation to where.

World Bank maps

Social Explorer maps

US Geological Survey's National Map

These show how measured paramaters change over time. The "shape" of the data provides additional information.

Temperature, precipitation, and snowfall in Marquette, MI

US import price indexes, 12-month percent change

Changes over time can also be shown spatially. For example, see this paper on Visual Storytelling in Cognitive and Geotemportal information Spaces.

Relational graphics are more abstract, comparing changes in 2 or more variables at a time.

Website graphs show the structure of web pages graphically

Parallel coordinate plots show many data points across multiple variables

Create a Graph from the National Center for Educational Statistics

Graphics can be misleading ... either intentionally or unintentionally. Problems arise when:

"Disappearing baseline" uses varying scales
Visual representation is inconsistent with numberical representation (as in this bubble chart)
Unexpected changes in scale appear
Change in height is interpreted as change in volume

Problems can arise because: