On procedure documentation: how is it changing the way that I work?
As a researcher it’s sometimes interesting to step back and ask myself “how did I get to this place in my project?”. Oftentimes in the middle of the research process I can give you a general idea- first I tried x, then I pivoted to y, and now I’m working on z ect. Whatever path I’ve taken through the research process oftentimes feels like it makes perfect sense to me, even if I can barely tell you why I decided to head down each road in the first place.
This process was tested this week as I set out to write my pre-research plan for an analysis of congressional gerrymandering in Alabama. Instead of my usual let-it-rip process I instead utilized HEGSRR’s open source template for reproducible geographic research to pre-plan my research approach. This included documenting data source metadata, recording processing environment and package metadata, and detailing data transformations before I ever hit ‘run’ on any code. This forced me to really think intentionally about what I wanted to get out of each piece of data. Additionally, I spent time dictating what different results would mean in context of the study in an attempt to discourage cherry-picking significant results.
This process was especially interesting to me because it forced me to put to paper (or vsc, in this case) what my thought process was. In a lot of ways this is what I’ve spent the most time developing over my four years at Middlebury; hard skills come and go, but what doesn’t is your ability to look over a dataset and make decisions about how to treat data. I’ll be curious to see how my workflow does during implementation, which should get done this week.