Spatial map
From Business Intelligence
 Interactive Map of the University of Oregon. This topographic map shows geographical features of the University of Oregon campus while also showing selected physical features. It makes good use of simple color coding to help the user intuitively interpret the map, and its interactive zooming features enhances navigability. However, the bright green used to represent the artificial turf field is probably extraneous since artificial turf is a seldom enough element not to warrant its own color, and because there is probably not a need for the average to distinguish between real and artificial turf. Consequently, it pulls attention away from the map to show information that is not necessarily useful. |
 UM: Commuter Southbound route map. This thematic map rejects geographical features and strict accuracy in favor of a representation that makes sense for its medium. A bus rider is not concerned with how much the road actually winds, but rather how to get from point A to point B. Highlighting the route, adding selected important data points, and using a monochromatic color scheme brings out relevant data well. There is a screenshot of an older version of the map here, and the improvement the highlighting makes is apparent. |
 The Shrinking Student Pool and Higher Education: An Example from Pennsylvania. This is a cartogram depicting Pennsylvania counties proportional to the number of students attending. It distorts county area according to enrollment, and additionally uses color coding to depict quantitative value. The color coding is redundant and the multiples do not need frames around them, but otherwise it is a good visualization for the data. |
 Violent incidents on US campuses. This is a country map with small multiples. Small multiples are useful for many types of visualizations, but sometimes more problematic for spatial maps. They require frequent back-and-forth visual comparisons to make sense of the data. Additionally, the range is not well set for this data set, as it is difficult to tell much of anything about campus crime in places other than Arizona. |
 Violent incidents on US campuses. This is the same data as before, except utilizing a dot map wherein the bubbles represent values, and the effect is overall more useful. |
Description
This method is used to show data in its spatial/geographical context. This is useful for comparing single variables within different areas, for instance, the average income in different US states. There are several maps that can be used for data visualization:
- World map: A chart that shows values on geographic regions, such as visually depicting categorical differences between different countries. ManyEyes supports depicting this via color changes, or via bubble sizes much like a bubble chart. Learn more about how to present data in a world map at ManyEyes.
- Country map: A chart that shows values on geographic regions, such as visually depicting categorical differences within a country's states, provinces or territories. Learn more about how to present data in a country map at ManyEyes.
- Topographic map: A map preserving geographic features while visually representing other features. This allows people to link area features with information. Examples include a map that shows subway stops, or some of the local business information available in Google Maps.
- Thematic map: A map with data superimposed on it. For instance, a map of the United States could show a pie chart for each state. This differs from a topographic map in that it shows more specific information. Learn more about the thematic map at Wikipedia, and look here about the difference between a thematic map and a topographic map.
- Cartogram: This map shows values using distorted land areas rather than variables such as color. Learn more about the cartogram at Wikipedia.
Considerations
- The problem with maps that use small multiples is that side-by-side comparisons can be difficult for people to parse, especially if many different variables are involved.
- For a map-generating program, the data set needs to contain standardized region names.