For our last exercise, we focused on the Modified Area Unit Problem (MAUP) and it's Scale and Zonation Effects related to spatial data analysis and display. The shape and size of vector data, such as polygons and polylines, can effect how they are interpreted. The “coefficients” or lengths, perimeters, polygon counts, and areas increase in value as the “number of areal units representing the data'' decreases. The problems occur in the "clustering" of sample points and the exclusion of data as detail is easier to see, but things become "left out."
In the last part of our exercise, the specifics of how these effects can cause problems are explored through the concept of gerrymandering, where voting districts or "zones" have strange shapes, sizes, and/or geographical decisions being made in a possible attempt to influence how votes are collected and/or counted. In the screenshot below, you can easily see how the polygon is potentially being manipulated with no obvious evidence for any significant need:
To find this, we used the perimeter-based Polsby-Popper test to score how "compact" the districts were, with the worst offender being a district located in North Carolina. North Carolina had more than one district that scored badly, while some districts made up entire states and scored much higher for the test. The test consists of a simple equation to compare the amount of land to a circle, with the results closer to "1" being better.
A conceptual lesson with valuable insight into the display and interpretation of spatial data! And, the importance of ensuring that data is being accurately analyzed for the best result, as well as communicated honestly and reliably for practical use. An excellent end to another rewarding semester at UWF!
Comments
Post a Comment