東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Multivariate ordinal responses in cl...

Gange, Stephen John.

Linked to FindBook

Google Book

Amazon

博客來

Multivariate ordinal responses in clinical trials.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Multivariate ordinal responses in clinical trials./
Author:	Gange, Stephen John.
Description:	176 p.
Notes:	Source: Dissertation Abstracts International, Volume: 55-05, Section: B, page: 1906.
Contained By:	Dissertation Abstracts International55-05B.
Subject:	Statistics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9419129

Multivariate ordinal responses in clinical trials.
Gange, Stephen John.

Multivariate ordinal responses in clinical trials. - 176 p.

Source: Dissertation Abstracts International, Volume: 55-05, Section: B, page: 1906.

Thesis (Ph.D.)--The University of Wisconsin - Madison, 1994.

Over the past few years, the biostatistical community has taken an increased interest in the analysis of correlated categorical data. Little research, however, has investigated how the statistical methodology developed for such data can be integrated with current clinical trial practices. The research discussed within this thesis investigates several aspects of medical studies, with an emphasis in ophthalmologic studies, that have correlated ordinal responses modelled as a function of regression coefficients.Subjects--Topical Terms:

517247
Statistics.

Multivariate ordinal responses in clinical trials.
LDR:03234nmm 2200313 4500 001 1840564
005 20050802071621.5
008 130614s1994 eng d
035 $a (UnM)AAI9419129
035 $a AAI9419129
040 $a UnM $c UnM
100 1 $a Gange, Stephen John. $3 1928894
245 1 0 $a Multivariate ordinal responses in clinical trials.
300 $a 176 p.
500 $a Source: Dissertation Abstracts International, Volume: 55-05, Section: B, page: 1906.
500 $a Supervisor: David L. DeMets.
502 $a Thesis (Ph.D.)--The University of Wisconsin - Madison, 1994.
520 $a Over the past few years, the biostatistical community has taken an increased interest in the analysis of correlated categorical data. Little research, however, has investigated how the statistical methodology developed for such data can be integrated with current clinical trial practices. The research discussed within this thesis investigates several aspects of medical studies, with an emphasis in ophthalmologic studies, that have correlated ordinal responses modelled as a function of regression coefficients.
520 $a After a review of the literature, a new marginal model for correlated ordinal data is described. This model is a generalization of a widely used model for univariate ordinal data, and estimation is based on generalized estimating equation methods. This new modelling strategy is compared to a current practice in ophthalmology of combining ordinal responses from two eyes into a single ordinal response, which I call a "composite response method." In particular, the relationship between regression parameters from the two models is formulated for both independent and correlated observations, and an example is provided using data from the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR).
520 $a For simulation studies and experimentation with the new ordinal regression model, a method of generating random responses based on a known set of model parameters was needed. I propose a new way of generating the random responses by reformulating the problem in terms of fitting a log-linear model to a contingency table.
520 $a A practice that is widely used within the clinical trial community is the use of interim monitoring techniques, which allow investigators to stop the trial before the prespecified endpoint through the use of sequential stopping procedures. Two critical issues in sequential monitoring the proposed model for correlated ordinal data are identified and investigated. The first concerns the underlying theoretical structure of the problem, and the applicability of existing statistical computer programs for analyzing this type of data. The other critical issue is how to best measure "how far along" the trial has progressed. Several suggestions are made and evaluated, and different methods are compared using data from the Early Treatment Diabetic Retinopathy Study (ETDRS).
590 $a School code: 0262.
650 4 $a Statistics. $3 517247
650 4 $a Health Sciences, Ophthalmology. $3 1019445
650 4 $a Biology, Biostatistics. $3 1018416
690 $a 0463
690 $a 0381
690 $a 0308
710 2 0 $a The University of Wisconsin - Madison. $3 626640
773 0 $t Dissertation Abstracts International $g 55-05B.
790 1 0 $a DeMets, David L., $e advisor
790 $a 0262
791 $a Ph.D.
792 $a 1994
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9419129