Student Seminar Series - June 13, 2006
University of Minnesota
School of Statistics
College of Liberal Arts

A Review of Quantile Regression with Application to Low Birth Weight Data


Marit Gilkey


Tuesday, June 13, 2006
1:00 PM, B29 Ford Hall
Minneapolis, East Bank Campus

Refreshments at 12:30 PM
300 Ford Hall


Abstract


For decades, ordinary least squares regression has been used to investigate the relationship between a response and the covariate(s). Quantile regression has not been around as long, but can give a more complete picture of the relationship between variables by modeling the conditional distribution of the response given the covariates at different quantiles. This project provides a short review of the theory and applications of quantile regression, including estimation of conditional quantiles, hypothesis testing, and model selection. Simulation results are provided to compare quantile regression with least squares regression. The advantage of quantile regression is observed in several situations. To illustrate the application, the quantile regression method is used to analyze a low birth weight data set.