This week, we looked at how various data classifications can visual the same data differently. We used census data from the Florida Geographic Data Library to create simple choropleth maps of seniors in Miami-Dade County. We visualized the percentage of people 65 and above using equal interval, quantile, standard deviation and natural breaks. Next we used population counts naturalized by square mile and created a similar map. I'm including the map that uses the percentage because it better visualizes the data.
My data classification was a little bit of a Goldilocks story! The equal interval method creates classes that are spaced equally along the entire range (0-79%), but presented the data in such a way that looks like very few seniors live in South Florida. Several of the classes don't even appear on the map because there aren't any records that fell into these. The Quantile method put equal numbers of records into the 5 classes ArcGIS computer, but this skews the data in the opposite direction - making it look like high numbers of seniors everywhere! The Standard Deviation method calculates how far the records are from the median, which in turn highlighted the higher and lower than average population areas. This just resulted in a map that was hard to read. So, we come to the just right - the Natural Break method, in which ArcGIS created the classes by lumping similar numbers together. This method allows the viewer to see more nuance in the data.
This has been my favorite assignment so far. I am not the best graphic designer and get frustrated with some of the layout details. I have been referring to them as the fiddling bits as you can sit and fiddle with them forever! I had more fun thinking with this one, thinking about how the classification methods visualized the data so differently and how this can change the meaning and implications of the data.
No comments:
Post a Comment