13: Visual Display of Quantitative Data
"Excellence in statistical graphics consists of complex ideas communicated with clarity, precision, and efficiency." – Edward Tufte.
"Graphics must not quote data out of context." - Edward Tufte.
"If you look after truth and goodness, beauty looks after herself" – Eric Gill.
Lecture outline: Best practices in presenting quantitative data visually.
Preamble: why use visualization
Preamble: clutter, the lurking enemy of visualization.
1. Presenting quantitative data in tables
Show data variation through design
Improving by providing context
Improving by rearranging/ rounding.
2. Statistical graphics: a history
Time-series graphics: William Playfair
Bar-charts: another invention by Playfair
A visualization success story: Florence Nightingale’s Coxcomb graph
Charles Minard’s ‘visually eloquent’ champion graph
Modern visualization developments: Tukey, Tufte and Cleveland.
3. Excellence in statistical graphics.
Tools for highlighting patterns:
Pre-attentive visual pop-out attributes
Visual perception tools
Graphical integrity:
Avoid confounding with poor design
Highlight uncertainty of data
Use connected lines for quantitative data only
Avoiding chart-junk (unnecessary chart embellishments that serves to distract)
Graphical sophistication:
Choosing the right graphing method
Provide context for data
Increase data-to-ink ratio; Avoid redundant ink.
Removing cognitive load
Highlight data variation through design
Layering to build visual hierarchy
Primary reference for this lecture:
“The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling” by Raj Jain; Chapter 10: “The Art of Data Presentation”.
Secondary references for this lecture:
1. “The Visual Display of Quantitative Data” by Edward Tufte
2. “The Joys of Stats” by Hans Rosling, Documentary on BBC Four [Link to an online version on gapminder.org].
3. “Head First Data Analysis” by Michael Milton.