Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1149117 | Journal of Statistical Planning and Inference | 2010 | 10 Pages |
Abstract
Suppose that we have a linear regression model Y=Xâ²Î²+ν0(X)ε with random error ε, where X is a random design variable and is observed completely, and Y is the response variable and some Y-values are missing at random (MAR). In this paper, based on the 'complete' data set for Y after inverse probability weighted imputation, we construct empirical likelihood statistics on EY and β which have the Ï2-type limiting distributions under some new conditions compared with Xue (2009). Our results broaden the applicable scope of the approach combined with Xue (2009).
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Applied Mathematics
Authors
Yongsong Qin, Qingzhu Lei,