Full Citation

Title: Predicting Elections with "Mister P" (MRP, multi-level regression and post-stratification)

Citation Type: Miscellaneous

Publication Year: 2016

ISBN:

ISSN:

DOI:

NSFID:

PMCID:

PMID:

Abstract: In this project I used Bayesian multi-level regression and post-stratification (MRP), and Python and [PyMC3](https://github.com/pymc-devs/pymc3) to adjust an unrepresentative sample of election preferences and predict the outcome of the 2016 US Presidential election. My results don't match the survey from which I used the raw data, possibly because I haven't used enough predictors (in the interest of simplicity). Nonetheless, the code and explanation show an attempt at learning and applying MRP using a language and MCMC framework that I haven't found used for such a purpose before.

Url: https://github.com/aenfield/Data512/tree/master/project

User Submitted?: No

Authors: Enfield, Andrew

Publisher: University of Washington

Data Collections: IPUMS CPS

Topics: Methodology and Data Collection, Other

Countries: