碩士 / 逢甲大學 / 統計與精算所 / 98 / We are concerned with the estimation problem of logistic regression model with incomplete response variable that is missing at random. We discuss three estimators that account for the missingness of the response variable.The proposed estimators are based on the inverse probability weighted estimating equations and are some extensions of the well-known Zhao and Lipsitz (1992)''s estimator.The first estimator assumes that the selection probablity is known, while the second and third estimators use a parametric estimation and a nonparametric estimation of the selection probability respectively.We develop the large sample properties of these estimators and compare their efficiencies through simulations using various values of the sample size, missing rate and correlation coefficient between the response variable and a surrogate for this variable.The methods are illustrated with real data from Pepe (1992) and Pepe et al. (1994).
| Identifer | oai:union.ndltd.org:TW/098FCU05336018 |
| Date | January 2010 |
| Creators | WEI-XUAN ZHANG, 張瑋玹 |
| Contributors | Shen-Ming Lee, 李燊銘 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 59 |
Page generated in 0.0056 seconds