Data Visualization (ChicagoRuby)

Data Visualization
Hosted by ChicagoRuby at Thoughtworks
Presented by Vanessa Shen (@vanessashen)

The size of the visual should match the value of the data

Distorting Data:
-3D can make data more complicated
-Pie charts are particularly difficult to understand with more than 3 values/slices; 3D makes pie charts even more difficult
-Avoid gradients unless the color variation means something; don’t use them just for decoration

Example of an easy to explore and compelling visualization: NY Times’ How Different Groups Spend Their Day –

Maximize data to ink ratio:
Remove ink that doesn’t covey info (decoration/chart junk) or is redundant (info repeated elsewhere in your presentation)

Word Clouds: Sideways or small words don’t get read

Use color
-Sequential (like gradients that represent percentages)
-Diverging (like political maps that combine two gradients, red to purple and purple to blue)
-Categorical (a set of colors where each one is very different from the others; different because you font want to imply a false relationship between any colors)

Keep in mind color relationships that are ingrained ; green to yellow to red might imply good to neutral to bad. (We are what we eat example)

Avoid hovers: if the info is important for the user to see it should be fit into the visualization

Chart junk: Fancy visualization does not equal good visualization
Bad fancy visualization: (USA Today example)

Good fancy visualization:
Watch the growth of Walmart and Sam’s Club by Flowingdata – displays change in quantity over time in a compelling way:
The Jobless Rate for People Like You by NY Times – actually a very simple visualization of complex information:

Data doesn’t have to be made interesting for all people, rather it should be made clear for people who will find it interesting anyway

Google Chart Tools – Automated chart making
D3.js or Processing – Charts from scratch
R – a statistical language, good for viewing and interpreting data but not necessarily a final product

Edward Tufte – The father of modern DV
NY Times – They’re good at this (see above. See also their ChartsNThings Tumblr:
Amanda Cox
Mike Bostock
The Why Axis
Fernanda Viegas and Martin Wattenberg at Google
Jeffrey Heer

Info graphics=chart junk

Hans Rosling’s Ted talks:

Also of note from the Q&A: not much tall in the data visualization space about mobile. Case in point, many of the about examples are flash and I can’t see them on my phone.