This week during my training, I practiced geographic data analysis using ArcGIS. I expect ArcGIS will be a useful tool for my classes and work in the future because I want to be a historian. It was important to me this week that I created both a dataset and a map. I wanted to learn more about how to format a spreadsheet of geographic data for ArcGIS. I wanted to graph places in my hometown, Bowling Green, OH. I wanted to show the places I’ve had the most fun in the 15 years I lived there. I listed the amount of fun I had in the place (1-10), the relative amount of risks I took in the place (1-10), the number of years I spent having fun in the place, and whether I spent time with friends, family, or both in the location. I created this dataset because I suspected there was a discrepancy between my nostalgic memories of fun times compared with the reality of where I had the most fun.
My data appeared as data points based on the addresses I listed. (In fact, I had fun creating the dataset because recalling the addresses of each location was a fulfilling challenge for me as a former U.S. Postal Service worker whose job required me to memorize the addresses in my town.) I added different layers to emphasize my fun, risk, and years variables. Then I customized how these points appeared so all of this information could be displayed at once. I experimented with the new Map Viewer and sketched neighborhoods and regions of my town. The pop-ups with my brief description of the area is not visible in Map Viewer Classic, but I was glad to try working with the new Map Viewer.
Then I ran my analysis. I used ArcGIS’s analysis tool to describe the distribution of my “fun” data. It created an eclipse showing an area two standard deviations from the center of my distribution. I was surprised how large the geographic area representing two standard deviations was. It included a variety of locations that I would absolutely not associate with any fun memories, like the Wood County Hospital and the spot where Pretty Boy Floyd shot a Bowling Green police officer in the 1930s. There were a few locations which, although they feature prominently in my memories like my high school, were simply too far from the downtown region. My map helped me realize just how geographically concentrated much of my fun experiences have been.
I was a lot more comfortable customizing my data visualizations this time because I knew which options were available. This time, I was limited by the size of my dataset. If I had had a few more hours to be more thorough and accurate I have no doubt I could have created a more revelatory map. I wish I could have found a way to add a very localized map of median household income or some indicator of poverty that could display differences at a level smaller than a census tract because our census tracts are so large that individual neighborhood-level differences aren’t visible. I wanted to see if there was a correlation between the fun I had, the risks I took, and the economic strength of the location. I also would have included a title for each location because my map currently has the addresses, rather than a description. The addresses are meaningful to me, my family, friends, and residents of Bowling Green, but they mean nothing to others.
ArcGIS’s functionality depends heavily on the quality of the data that you upload. I was glad to get some practice with thinking about how to create variables that could be displayed through mapping. This week’s training was an entertaining way for me to begin thinking more deeply about digital geographic analysis.
