soodoku / birthday_voter

Hi there 👋

birthday_voter's Issues

data and measurement section

data and measurement section needs:

1. summary table with:
   year, election_type, election_date, n_eligible_voters, prop_voted

2. for the data validation checks, it is not enough to say X passed. it is good to describe the results.

   a. waffle plot for the first point though density plot may be interesting also. talk about the max and min and why they are reasonable.

   b. loess b/w age by prop_voting

   b. we can say on average midterm was higher than primary by X%. ideally, we want to refer to a table. same for primaries

3. details on how much we lose when we filter and a line on why we filter may be useful.

4. how 'voted' was coded. in the appendix, we will have to show results for different versions of voted.

change output of tables from png to csv

change table and figure names to semantic + numbering

change file names for figures and table to a format like: fig1_voting_percentage_by_age.pdf

OH, PA, and WA analysis

analyze by race

reproduce tab 1, 2, and 3 and figures by race

download and clean the data

1. Download FL voting reg. data here:

https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/UBIG3F

there are more states but let's start here.

2. Clean data:

some people use 1/1 etc. as fake b'day. investigate a bit and then let's remove those rows.

3. then the third step is simply checking the proportion of people who vote when election day in that year = their birthday vs. people with birthdays on any other day but especially with birthdays a day before or after the election day.