USING DIAGNOSTIC MEASURES TO DETECT NON-MCAR PROCESSES IN LINEAR REGRESSION MODELS WITH MISSING COVARIATES

Authors

  • Shalabh .
  • Helge Toutenburg
  • Andreas Fieger

Keywords:

Linear regression model, missing data, missing completely at random, diagnostic.

Abstract

This paper considers the problem of missing data in a linear regression model.
It presents a method to analyze and detect the missing completely at random
(MCAR) process when some values of covariates are missing but corresponding
values of response variable are available. The idea of using outlier detection
method in linear regression model is proposed to be employed to detect a non-
MCAR processes. Such an idea is utilized and a graphical method is proposed to
visualize the problem.

Downloads

Issue

Vol. 44 No. 2 (2010)

Section

Articles