Monte Carlo Simulation Study of Biased Estimators in the Linear Regression Models with Correlated or Heteroscedastic Errors


Ozkale M. R.

COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, vol.43, no.5, pp.1143-1186, 2014 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 43 Issue: 5
  • Publication Date: 2014
  • Doi Number: 10.1080/03610918.2012.728273
  • Title of Journal : COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
  • Page Numbers: pp.1143-1186

Abstract

In this study, the performance of the estimators proposed in the presence of multicollinearity in the linear regression model with heteroscedastic or correlated or both error terms is investigated under the matrix mean square error criterion. Structures of the autocorrelated error terms are given and a Monte Carlo simulation study is conducted to examine the relative efficiency of the estimators against each other.