The primary objective of our lab for week six was to experience the differences between classification methods in how our data are presented. The first map above presents the same data filtered through four different classification methods. None of them lie, but all tell only versions of the truth. Being aware of how one's chosen classification method affects data presentation and audience perception is crucial in understanding how to communicate your analysis.
As can be seen above, the four classification methods focused on for this lab were: Natural breaks, equal interval, quartile and standard deviation. The equal interval classification scheme is determined by taking the data range (from the highest to the lowest sample value) and dividing it by how many classes one has chosen (e.g., five in the example above). This produces a classification that is not responsive to the data; some categories may have the bulk of samples while others may have few or even none. The quartile method is, in a sense, the opposite of equal interval by creating classes such that each have the same number of samples. Standard deviation, as the title suggests, creates categories based upon statistical standard deviations. This method, while certainly useful for some people, is more difficult to interpret than quartile and equal interval.
The second map above focuses on the natural breaks classification method. This method was designed to create classes based upon how the data are grouped; ideally, each class represents a data group and class breaks do not divide groups. As noted in our lab instructions, this is ArcMap's default classification method and, in my opinion, most clearly communicates the given data.
No comments:
Post a Comment