What you should know about Classification Methods with Graduated Colors/Symbology
I wanted to take a minute here to answer a question I initially had when exploring map design, specifically with graduated colors and symbology: what do the different classification systems mean? When creating any type of choropleth map, understanding the classification methods (Equal Interval, Defined Interval, Quantile, Natural Breaks/Jenks, Geometrical Interval, Standard Deviation) is essential to understanding your data. Depending on the different method you use, the exact same data can convey a VERY different message across the map. Here are just a few quick things to remember when choosing a classification method for your graduated color scheme or symbology:
Obviously, you can manually input your own breaks if you have a specific range you are trying to interpret, but most of the time we might want the computer to do some of the work, but what is Arc doing with your data when you make a selection? The three you should really get a firm understanding of are equal interval, quantile and natural breaks or Jenks.
The equal interval classification method is exactly what it sounds like; it divides the feature data into equal-sized subranges of the range available. So say your data ranges from 1 to 90 and you designate 3 intervals, the software will create the intervals of 1-30, 31-60, and 61-90. This type of classification works best when you have data that is normally distributed.
However, if you have a data set with many outliers and or is linearly distributed a better choice might be the quantile classification method. This classification method divides your data into equal amounts of features within each interval.
The other classification method that is important to understand are natural breaks or Jenks. Natural breaks are determined by the software through a built-in algorithm that is too complex for me to understand. And after briefly looking into what the actual algorithm is, I found out that it is "proprietary information" or top secret Esri business, so I guess we will never know. But, what natural breaks essentially do is, based on the data within the feature class, determine where "natural" breaks in the data fall. From there, you could go in and edit the breaks manually to round the break up to more intuitive number.
Image source
Clearly, there are a lot of ways you can break up your data in ArcGIS with classification systems, and which ones you use are based on personal preference and also the research question you are trying to answer. As you can see in the image above, the visualization you get with each type of classification method is very, very different depending on the one selected. None of the methods can legitimately be classified as "good" or "bad" but understanding the differences in how the breaks are sorting the data displayed on the map is a crucial step in interpreting the correct patterns going forward.