I’m doing some form of data visualization learning for 180 days because I need to #JFDI.
Week Four Discussion Question
Geographic maps are undoubtedly useful tools for communicating data visually, yet they inevitably introduce distortions and mistruths. What are some of the compromises
that must be made when producing a geographic map? (Consider sharing an example to illustrate your point.)
The biggest example I can think of that pops off my head are maps that map data characters of populations over the map of the United States… when you just end up getting a map of the population distribution of the United States rather than being able to denote a trend. Also, regional-level maps also distort complexity.
For example, an electoral map by state doesn’t reveal the story of urban voting trends. On the flipside, the newly sectioned map does not account for population clusters.
Week Four Quiz
- D3 projections take a two-value array: longitude and latitude and return x and y as output
- You should value maps to area of each circle rather than radius to avoid distortion
Reading Nathan Yau’s book Visualize This
258-272 Spotting Differences
- Multidimensional Scaling can help you find clusters – there are various models you can use to do this p 261
- Ways to find outliers: bar charts, histograms, boxplots p 269
- When working with Maps data, think of x and y coordinates as longs and lats p 272