r-d Class Estimator Under Misspecification


Siray G.

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS, cilt.44, ss.4742-4756, 2015 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 44 Konu: 22
  • Basım Tarihi: 2015
  • Doi Numarası: 10.1080/03610926.2013.835421
  • Dergi Adı: COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
  • Sayfa Sayısı: ss.4742-4756

Özet

Omission of some relevant explanatory variables and multicollinearity in regression models are very serious problems in applied works. There are some papers examining the multicollinearity and misspecification which is due to omission of some relevant explanatory variables, concurrently. To remedy the problem of multicollinearity, Kacranlar and Sakallolu (2001) proposed the r-d class estimator that includes the ordinary least squares, principal components regression, and Liu estimators as special cases. The aim of this paper is to examine the performance of the r-d class estimator in misspecificied linear models.