For March, I originally set out to build an R visualization that runs on a production script.

However, two months in, I have realized that my focus is not on productionalization, but the quality of the findings that come from the analysis. With this in mind, I will use the same data set as last month and find deeper insights than simply publishing a cross tab.

Potential analyses include:

• Compare voting behavior in 2016 versus 2018 (segment by youth vote, senior vote, high income vote, etc)
• Cohort voting behavior as a time series (plot each group along life cycle)
• Predict 2020 presidential result in Ohio, using history data
• Identify voters that could be activated in 2020 (i.e. voted in 2016 but not in 2018)

I will work through these ideas and select a topic shortly.

Note: there is an opportunity to run an R script automatically, but since my voting history dataset does not change daily, this is not a priority at this time.