Online Program Home
  My Program

Abstract Details

Activity Number: 162 - SPEED: Government Statistics, Health Policy, and Marketing
Type: Contributed
Date/Time: Monday, July 31, 2017 : 10:30 AM to 12:20 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #323074 View Presentation
Title: Analysis of the Australian Election Study Using Bayesian Quantile Regression Models
Author(s): Charles Au* and S. T. Boris Choy
Companies: University of Sydney and University of Sydney
Keywords: Markov chain Monte Carlo ; quantile regression ; discrete choice models ; Bayesian inference ; Australian Election Study (AES)

Quantile regression models are useful for modeling the conditional quantile of the dependent variable. Recently, quantile regression models have also been applied to discrete choice data, where the dependent variable is binary or ordinal. Quantile regression is a non-parametric model. However, in the Bayesian approach, parameter estimation for quantile regression can be done via a parametric method, when the error terms of the model are assumed to follow, for example, the asymmetric Laplace distribution. This paper proposes the application of Bayesian quantile regression models to survey data from the Australian Election Study (AES). The binary and ordinal quantile regression models will be used for investigating the factors that influence voters' choice for certain political parties and their interest in politics. The main objectives are to investigate the differences in the coefficients estimates and model performance of the regressions at various quantile levels. Comparisons will also be made to binary and ordinal probit models.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

Copyright © American Statistical Association