Analysis Procedures
Our team conducted two major analysis procedures on our collected data in order to answer our guiding research questions. The procedures we used included both a descriptive statistics analysis for our data variables in the Promise Zone and an ArcGIS hot spot analysis of housing conditions in the area.
Descriptive Statistics Analysis for the North Hartford Promise Zone–
In this step, our team utilized basic statistical methods to find results from our datasets. This technique was useful for interpreting our economic background data, health data, and portions of our housing data. Our goal was to use simple measurements of the data like average and the concept of minimum and maximum. Then, we gathered measurements for the Promise Zone. We typically compared those measurements to measurements for the rest of Hartford. Our goal was to see where the Promise Zone stood out and for which variables. Here are a few examples of these descriptive statistics techniques. First, for example, we used the average or mean as a tool to compare the Promise Zone with the rest of Hartford on the question of sleeplessness. We calculated the average percentage of residents in the Promise Zone who get less than seven hours of sleep and then compared that number to the average for Hartford. Another example of our use of descriptive statistics was that we used measurements of dispersion (like minimum and maximum). For our analysis of asthma rates, for example, our team found the maximum and minimum asthma rates for the whole city of Hartford by tract. We then plotted each tract’s asthma rate for the city in order to discover where the Promise Zone’s ten tracts ranked within the city. These are just a few examples of what we have done. More examples will become apparent as you peruse the project story maps.
ArcGIS Hotspot Analysis of Housing Conditions-
For this procedure, we conducted ArcGIS’s “Optimized Hot Spot Analysis” for both housing blight and vacancy in the Promise Zone. The goal of a hot spot analysis is to determine where there are statistically significant concentrations of events (i.e. blight scores, instances of vacancy). To find these hot spots, we first had to upload the Community Solutions Property Conditions Survey into the ArcGIS software. We carried out a procedure called “geocoding” which allowed us to display each property as a point on a map of Hartford. After displaying each property, we then had to define the areas for which we wanted the program to conduct the hot spot analysis. To accomplish this, we had to draw the boundary of the North Hartford Promise Zone, being sure not to include areas without residential properties (i.e. Keney Park). For example, we did not want the program to look for hot spots across the entirety of Hartford or its parks when the property data we have is for the Promise Zone. Having determined these boundaries, we then ran the Optimized Hot Spot Analysis and were given the areas which were identified as having statistically significant clusters of both high and low blight score values in addition to instances of vacancy. In order to display these hot spots on the map clearly, we converted these clusters of points into a “raster” surface (surface of colored pixels) which allowed for spots to be displayed.