I can’t believe there’s only two weeks left of FURSCA. As the research portion of my summer starts to come to a close, I’ve been reviewing the work I’ve completed thus far.
I started FURSCA this summer with three goals – to identify the profile of a voter for the city of Lansing, to identify the profile of a protestor for Lansing, and to see if I could correlate voting and protesting within the city. I planned to use Census data and Facebook protesting data to achieve these goals, analyzing the data in SPSS.
The first two weeks were spent reading four different books – We The People, The Politics of Power, Introduction to Linear Regression Analysis, and the Michigan and United States sections of The Statesman’s Yearbook – 2011. This reading allowed me to have a better understanding of voting and protesting as it has existed and evolved in the United States, dating to pre-revolutionary times. This reading also allowed me to understand what SPSS (or Statistical Package for the Social Sciences) would do as it analyzed my data.
The third and fourth weeks were spent mainly in Lansing and Detroit, collecting data, information, and perspectives needed to complete my research. Lansing City Clerk Chris Swope was instrumental in getting me voting data – as he provided me with over 90,000 data points that I was then able to clean and organize into usable data for my FURSCA project. The Census Bureau office in Detroit provided perspectives on the national voting “scene”, which allowed me to later correlate my Lansing data to the national averages. The Capitol library in Lansing helped me get data and information on Lansing that would help me predict what the voter and protestor profiles should look like.
The fifth and six weeks of FURSCA were perhaps the most crucial to my research project. As I began to clean out the data, I noticed variables listed in Mr. Swope’s table that I couldn’t identify. Through conversations with him, I was able to identify them and further sort the data by the variables I had outlined in my FURSCA proposal. Through doing research on the Lansing area school districts, I was able to sort my data not just by age, gender, or location, but education as well. Additionally, through Senate and House district numbers, I was able to get a picture of what each voter’s elected official’s views were and see if that had any influence on whether people chose to vote or not.
After I had identified all the variables listed in the data, I began to run reports. I was able to get the profile of a voter and a registered non-voter for the city of Lansing, as well as a rough profile of a protestor (See attachment 1). I was not able to correlate voting and protesting within the city, as there was nearly 100x as much voter data as there was protestor data. I was able to correlate the voter and registered non-voter data for Lansing to national data.
As I was running these reports, I realized two very important variables I had not considered. As I was predicting what my voter would look like, I was considering what I thought to be the “population” of Lansing. However, because of Capitol Hill, my “population” thought was skewed. The majority of the people working in the Capitol or in the Congressional office are commuters. Additionally, many people that protest in Lansing aren’t actually from Lansing. Because Lansing is Michigan’s capital, many people come from all over the state to protest. However, in spite of these variables, I was still able to find conclusive data about the residential population of Lansing in terms of voting.

The data to the left is data for a registered voter (whether they voted or not). To input data into SPSS, it had to be numerical, so I changed the actual words for many of these categories into numbers. For example, under gender, female became 1 and male became two.
The State House and Senate codes indicate what Michigan legislative district the registered voter lives in, and the US Congress code does the same but on a national level. This allowed me to gain a picture of what area of Lansing the registered voter lives in.
The school code indicates what school district the registered voter lives in. There were four school districts represented in the data I had and I ranked them from 1-4 (one being best, four being worst) based on the school districts’ data.
Average age is computed using the descriptive statistics function instead of the frequency function, so I didn’t include that table here.
PROFILE:
Average Age: 47
Gender: Female
State House Preference: Democratic
State Senate Preference: Democratic
US Congress Preference: Republican
School District: Mainly from Lansing Public Schools area