I remember sitting in a newsroom in Louisville, Kentucky, plotting my next move. I had worked at the Courier Journal for more than two years, and as a result of a number of position changes, my far proximity to home, and the nature of the journalism industry itself, I was ready for something different. My mind immediately went to something I had always wanted to accomplish but had put on the back-burner: the completion of my master’s degree.
I reached out to the University of Arizona School of Journalism, where I had left roughly four years earlier just a project or thesis shy of a degree. I turned in my application and was accepted shortly after that. After simultaneously securing a job at the Arizona Daily Star, I was back at the same school in the same city I started almost a decade ago.
To say this project was a long-time coming would be an understatement.
I faced a number of challenges along the way, but chief among them was working with a pair of programs that were essentially in a foreign language to me. When I started this project, I didn’t know what a “block file” was. But thanks to help from many people along the way, I quickly learned they were a type of population and demographic data that was broken down by census block. I later learned where to find them.
The other big challenge I ran into was figuring out a way to do what I wanted to do with the resources I had. In a perfect world, I would have built a program from scratch that would have done exactly what I wanted. But for this project, I had to figure out a way to juggle a bunch of misfit pieces and turn them into something suitable for my project. Both programs I used relied on older data, from the 2012 and 2016 elections. I wanted to use the most-recent data from the 2018 election to illustrate potential changes to districts. I needed to track down a shape file of those results to import into the programs. I quickly learned the state of Arizona does not keep data in that file format. I had two options – reach out to a few groups that process the election data into those files, or build the files myself. Because of time constraints I opted for both. As I was in the middle of receiving help to build the data file, I got word that the Harvard Dataverse had come through with a shapefile. That came about two weeks before my deadline to turn in my final draft of my project.
It’s never easy doing something you’ve never done before. It’s even harder to do it without much technical knowledge of what you’re working on. And it’s extremely hard to do something academic when you’ve been away from that world for so long.
Throughout this whole journey, I’ve felt a lot like I would take one step forward and then two steps back. But ultimately, I got it done. And I produced something I’m truly proud of.