- Bib ID:
- 3283163
- Format:
- Book and Microform
- Author:
- Forbes, Andrew Benjamin
- Description:
- 251 p.
- Summary:
-
The odds ratio is a measure used in both prospective and retrospective epidemiological studies to quantify the association between a binary risk factor for a particular disease and occurrence of the disease. In many applications involving matched data, a log odds ratio regression model is used to express the dependence of the odds ratio on covariates thought to influence the exposure-disease association. The standard technique for estimating the regression parameters based on this model is conditional maximum likelihood estimation (CMLE). We initially determine conditions under which consistency and asymptotic normality of the CMLE are valid for a sequence of sparse 2 x 2 tables, and then extend these results to the general log odds ratio regression model. Focusing attention on case-control studies, we then consider the effect of measurement error in the covariates on the CMLE of the odds ratio, both asymptotically and in small samples.
We propose improvements of the CMLE and evaluate their performance.
- Notes:
-
- (UnM)AAI9027000
- Source: Dissertation Abstracts International, Volume: 51-04, Section: B, page: 1908.
- Thesis (Ph.D.)--Cornell University, 1990.
- Reproduction:
- Microfiche. Ann Arbor, Mich.: University Microfilms International.
- Subject:
- Statistics
- Other authors/contributors:
- Cornell University
- Copyright:
-
In Copyright
Contact us for information about copying.
Copyright status was determined using the following information:
- Material type:
- Literary, dramatic or musical work
- Published status:
- Unpublished
- Creation date:
- 1990
Copyright status may not be correct if data in the record is incomplete or inaccurate.
Other access conditions may also apply. For more information please see:
Copyright in library collections.
Request this item
Request this item to view in the Library’s reading room.