Masters Theses
Abstract
"In disciplines such as economics, social science, and epidemiology, estimating regression models using ordinary least squares can lead to biased estimates due to one or more of the independent variables being correlated with the error term. When such correlation is not ruled out, the covariate under suspicion is called endogenous, and the ordinary least squares estimate of the covariate effect is typically biased and yields inconsistent estimates of the causal parameter.
This thesis is a simulation study of the bias due to endogeneity, and how much of this bias is removed is determined by resorting to the 2SLS technique. The simulation study was carried out to investigate the bias under several different model assumptions, and the results are compared"--Abstract, page iii.
Advisor(s)
Samaranayake, V. A.
Committee Member(s)
Le, Vy Khoi
Bryant, Richard
Department(s)
Mathematics and Statistics
Degree Name
M.S. in Mathematics and Statistics
Publisher
University of Missouri--Rolla
Publication Date
Spring 2006
Pagination
vii, 56 pages
Note about bibliography
Includes bibliographical references (pages 53-54)
Rights
© 2006 Xujun Wang, All rights reserved.
Document Type
Thesis - Restricted Access
File Type
text
Language
English
Subject Headings
Equations, Simultaneous -- Numerical solutionsInstrumental variables (Statistics)Least squares
Thesis Number
T 8955
Print OCLC #
82371529
Recommended Citation
Wang, Xujun, "Endogeneity bias and two-stage least squares: a simulation study" (2006). Masters Theses. 3859.
https://scholarsmine.mst.edu/masters_theses/3859
Share My Thesis If you are the author of this work and would like to grant permission to make it openly accessible to all, please click the button above.
