Date of Graduation

8-2019

Document Type

Thesis

Degree Name

Master of Science in Statistics and Analytics (MS)

Degree Level

Graduate

Department

Graduate School

Advisor

Giovanni Petris

Committee Member

Avishek Chakraborty

Second Committee Member

John R. Tipton

Abstract

Our goal is to create spatio-temporal models for predicting future gubernatorial elections. For a concrete example of how well our models work we use past data to predict the 2018 Arkansas gubernatorial election and use the existing 2018 election data to check our models predictive accuracy. Gubernatorial election data was collected from the Arkansas Secretary of State website while related covariate data was collected from the website for the Federal Reserve Bank of St. Louis. The data we collect is on the county level. For predictive purposes we fit multiple models to the data using Markov chain Monte Carlo and compare each model to determine which has the best predictive ability.

Share

COinS