Different Map Classification Methods

 This week we learned about the different map data classification methods: equal intervals, quantile, standard deviation, and natural breaks. Utilizing these different methods, I analyzed how these methods highlight and possible skew different aspects of data while also comparing these same methods to normalized data tracts by area. 

1. Equal Interval: In this classification method, you divide your max value by your desired number of categories. When analyzing our data, this method causes a huge skew in the data where most all of the data landed in the first category.

2. Quantile: This method looks to create an equal number of observations within the number of desired categories. This method helped to display the data in the most diverse way possible.

3. Standard Deviation: This method uses the bell curve and standard deviation to create 5 categories with +/- .5, 1.5, >/< 1.5 standard deviation from the mean. This method helps to visually see how the data various from the mean in one direction or the other. This method demonstrated that there was a pretty good bell curve with not much data skewed either way.

4. Natural Break: this method seeks to group together like values to accentuate the inter-class variance. This method closely represented the Quantile method as well, but this method did not have as much observations in the upper range, with most of the data in the lower two ranges. This showed that most of the population fell beneath the 12.9% category.

"Which classification method do you think best displays the data for an audience looking to target the senior citizen population? Explain why."

I would recommend the Natural Breaks method, as this shows the biggest natural differences/jumps between percentage groups. This will help them to accurately pinpoint areas with high levels of seniors.

Figure 1: Senior Population Distribution of Seniors in Miami Dade County, Florida by percent of age 65+. 

Figure 2: Senior Population Distribution of Seniors in Miami Dade County, Florida by normalized area tracts, per sq mile. 

"If you were presenting data to the Miami Dade County Commissioners regarding the distribution of senior citizens, which data presentation do you think more accurately depicts the distribution?—the percent above 65 or the population count normalized by area?"

I would recommend the normalized data by area, because without actively comparing the percentage to the size it covers, one makes a decision without being fully informed. For example, 1% of 10 sq miles is a much greater number than 5% of 1000 sq miles. If you were looking at just the percentages in the percent above 65 models, you would think 5% is greater.