碩士 / 東海大學 / 統計學系 / 85 / In this article, we propose a method of adjusting for item
nonresponse based on a multivariate logistic regression
model. The probabilities of response
are estimated using a working likelihood approach. The
estimated probabilites are used in a Horvitz-Thompson type
estimation to reduse bias in the estimation of sample means for
every single item and sample correlation coefficients of
correlated items. Simulation results suggest that the method
reduces the mean squared error of estimation when the response
probabilities have been correctly specified by our model.
| Identifer | oai:union.ndltd.org:TW/085THU00337004 |
| Date | January 1997 |
| Creators | Su, Yi Show, 蘇儀修 |
| Contributors | Shen Pao-Sheng, 沈葆聖 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 15 |
Page generated in 0.0021 seconds