Testing biasedness of estimating equations in weighted regression with missing covariate data /
To obtain consistent estimators of regression ity Microfilm Inc. coefficients in weighted regression with partially missing covariate data, asymptotic unbiasedness of weighted estimating equations is required. We propose two statistical tests for biasedness of estimating equations in weighted parame...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1998.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=733050331&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | To obtain consistent estimators of regression ity Microfilm Inc. coefficients in weighted regression with partially missing covariate data, asymptotic unbiasedness of weighted estimating equations is required. We propose two statistical tests for biasedness of estimating equations in weighted parametric and semiparametric regression settings. The asymptotic null distributions of these two tests are derived under the assumption that the massiveness mechanism of the covariate is missing at random (MAR). A limited simulation study is conducted to investigate the empirical performance of both the parametric and the semiparametric tests. Resulting empirical levels and sizes are presented for these two tests. Furthermore, the effect of model misidentification for the selection probability function on both tests is considered. Some recommendations are also provided. |
|---|---|
| Item Description: | Vita. "Major Subject: Statistics". |
| Physical Description: | xi, 95 leaves : illustrations ; 28 cm. |
| Bibliography: | Includes bibliographical references (leaves 92-94). |