predictors_of_poling_errors

In this repository you will find code and data related to the resarch project Political Predictors of Polling Errors. The aim of this project is to develop a contextual understanding of polling errors and their triggers. Unlike most previous studies, we take a cross-election comparative perspective and put the theoretical focus on characteristics of the electoral contest which may encourage polling errors.

This repository is structured as follows:

• All data files are in the data folder. There is one subfolder for each election type investigated.

• The us_president folder contains code for

  • scraping US presidential polls from 2000 to 2016 in the scrape folder
  • adding covariates in the covariates folder
  • a hierarchical bayesian model for analysing us presidential polls in the analysis folder.

• The us_senate folder contains code for:

  • scraping US senate polls from 1998 to 2020 in the scrape folder
  • analysis of US Senate candidate characteristics cen be found in the candidate_characteristics folder
    • cleaning the scraped polls and merging additional covariates in the covariates folder
    • a hierarchical bayesian model for analysing us senate polls in the analysis folder.
  • analysis of Trumpists can be found in the trumpists folder

• The german_bundestag folder contains code for:

  • scraping German Bundestag electiopn polls from 1994 to 2021 in the scrape folder.
  • There are subfolders in the analysis folder for:
    • data preparation in the preparation folder,
    • descriptive plots in the desc folder,
    • fitting a hiearchical bayesian model for analysing German Bundestag polls desc folder,
    • visulaizing results in the results_vis folder.

• The german_landtag folder contains code for:

  • scraping German Landtag election polls from 1994 to 2021 in the scrape folder
  • There are subfolders in the analysis folder for:
    • data preparation in the preparation folder,
    • fitting a hiearchical bayesian model for analysing German Landtag polls desc folder,
    • visulaizing results in the results_vis folder.

Detailed descriptions for all code files can be found in the respective folders.