Interpenetrating subsampling regression estimation with or without jackknifing

Abstract

In this paper four regression estimators are considered for a finite population total based on interpenetrating subsamples, two of which are with jackknifing and the other two are without jackknifing. Both theoretical and empirical comparisons of the four proposed estimators are done with respect to bias, variance and mean square error.

Keywords:

• bias
• interpenetrating subsampling
• jackknife
• mean square error
• regression estmation
• simple random sample
• variance