Title: Improved shrinkage testimators for the simple linear regression model

Authors: Z.A. Al-Hemyari; I.H. Hussain

Addresses: Department of Mathematical and Physical Sciences, University of Nizwa, P.O. Box 33, Postal Code 616, Oman. ' Department of Statistics, Al-Mustansiryah University, Baghdad, Altalibya, Iraq

Abstract: The performance of an estimator can be improved by incorporating some additional information available, besides the sample information. Our interest is to estimate the mean, slope, and response of simple linear regression model when some prior guess values are available from the past. The relative performances of the proposed estimators are investigated based on bias ratio and relative efficiency. It is observed that the proposed estimators fare better than classical estimators and existing shrinkage estimators.

Keywords: regression line; shrinkage; significance levels; bias; relative efficiency; linear regression modelling; estimator performance; estimators; testimators.

DOI: 10.1504/IJIDS.2012.045129

International Journal of Information and Decision Sciences, 2012 Vol.4 No.1, pp.87 - 103

Published online: 09 Aug 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article