Module 6 Data Classification - At least I'm moving in a positive direction!

This lab was designed to provide practice in not only using data classification methods in ArcDesktop, but also to illustrate the differences between the various methods.  We used Equal Interval, Quantile, Natural Breaks and Standard Deviation in this lab.  Each method needed to be properly symbolized by quantity so that the results were clear to the end user.  All used a color ramp, except for standard deviation which used a divergent color map so that the diverging values above and below the mean were distinguishable.  This lab continued to emphasize the importance of cartographic design to make the information easily interpreted by the end user.  As the map production process evolved, the application of good cartographic design steadily improved the appearance and usability of the map.

I feel the Map presenting the quantity of seniors per square mile is the most accurate means to portray the data.  The population count normalized by area more accurately depicts the distribution of senior citizens because only the senior age group is taken into account with the square mile area, not other age groups.  With this map, I compared the following classification methods:

·        With the Equal Interval method, the range is divided into equal parts along a number line and the data falls within the resulting classes based on its value.  This method did not depict the localized areas of seniors as well because the data was not value grouped.

·        Using the Quantile method, data is rank ordered and an equal number of observations are placed in each class.  Differentiation between data clusters is more visible here, but the high end of the ranges was greatly affected by the outlier.   

·        The Standard Deviation method considers how data is distributed along a number line; classes are created by adding or subtracting the standard deviation from the mean.  This method works very well with normally distributed data.  There is visual evidence of cluster differentiation and the effect of the outlier is visually minimized.

·        In the Natural Break method, similar value data is grouped and algorithms are used to minimize the value difference within classes and maximize the value differences between classes.  I preferred this method above the others as it made the population densities easy to distinguish and reduced the significance of the outlier on the overall data.

This lab certainly helped to make sense of the classification methods.  Combining this with proper symbology really helped to present the information in a clear, easy to understand format.