Cluster Analysis

Problem

In this exercise there were two scenarios that required two different types of analysis. In the first, the fire department would like to know if there is a relationship between the number of calls and income level. In the second scenario the county economic development office would like to know where there are clusters of high income and low-income homes so that they can direct their charity-collection efforts (high income areas) and job creation efforts (low income).

Analysis Procedures

  • Strategies

To address the problem put forth by my client, ESRIs ArcMap was used to perform the cluster analysis. Tools from the analyzing patterns tool set were employed including the Cluster and Outlier Analysis, and the Hot Spot Analysis. The data used for the Fire Station’s analysis included (fire station) battalion boundary layers, EMS Call data for Jan 2015, existing and proposed station points, and census data on median household income. For the Dallas Count economic development office’s analysis the county data including major roads, and the (2010) census data for household incomes were used.

  • Methods

Cluster & Outlier Analysis

I used the Cluster & Outlier Analysis tool inputting the call data from Jan 2015, selecting FEE and Inverse distance as the spatial relationship conceptualization method. I then copied the output layer and symbolized it so that HH values were one color, HL & LH values were another and LL yet another. Then I resymbolized the census data using graduated colors to get a range of values for each census block that give a visual representation of the median income level in each block. I then ran the same Cluster & Outlier Analysis tool with the same inputs except that I used a fixed distance band of 900 ft to conceptualize the spatial relationship. I then copied and resymbolized the results as I did in the previous analysis.

Hot Spot Analysis

I input the census block data in the Hot Spot Analysis Tool selecting the median income field (in this case P053001) and a distance of 5280 ft. The resulting output allowed me to make a map depicting where there are clusters of high and low income census blocks.

Results

Application & Reflection

In this assignment, I learned how to do a couple more types of cluster analyses. Other applications for this type of analysis would include looking for clusters of different types of crimes. In the case of the hot spot analysis you could even use the census data as I did in this exercise to look for patterns in what could be causing high concentrations of certain types of crimes. For example you could investigate the connection between population density and property crimes or income levels and violent crimes. You would need to obtain the crime data however from the local police district.