EJ Arce
Jonathan Matz
We have obtained multiple datasets of results from California state elections from 1968 through 2016. The data include voting outcomes for state-level offices, such as state governor, as well as state legislative offices. On a national level, the data also include the voting outcomes for presidential and US Senate offices. We will clean and merge our data into two datasets: one for state legislative office voting, and one for presidential, gubernatorial, and US Senate seat voting. We will then analyze the outcomes for each election, and model around such outcomes by testing possible covariates that could influence partisan voting patters. We will gather our own covariate data that we believe play a factor. For example, incumbency, country-level and state-level economic data, partisanship, and government spending could each influence how California citizens vote in a given election. Once we have gathered data on our covariates, we will test and develop the most accurate model to predict 2018 mid-term results.
- April 17: Legislative election results, national offices results, and covariate data all clean and joined for further modeling
- April 19: Initial modelling and 2+ visualizations
- April 24: Covariates and model improved
- May 6: Deliverables finished
- May 9: Model finalized, 2018 simulation run and final visualizations generated
- May 10: Poster created and project presented
Data and guidance provided on this project by Dr. Carl Klarner.
This repo is organized following the guidelines outlined here, adapted for use in this class from work by Project TIER and Thomas Leeper
— Jay Lee, Reed College