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

Fitting and testing continuation rat...

The University of Toledo.

Linked to FindBook

Google Book

Amazon

博客來

Fitting and testing continuation ratio logistic regression models based on case-control data.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Fitting and testing continuation ratio logistic regression models based on case-control data./
Author:	Peng, Cheng.
Description:	116 p.
Notes:	Adviser: Biao Zhang.
Contained By:	Dissertation Abstracts International64-07B.
Subject:	Statistics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3097835
ISBN:	9780496454488

Fitting and testing continuation ratio logistic regression models based on case-control data.
Peng, Cheng.

Fitting and testing continuation ratio logistic regression models based on case-control data. - 116 p.

Adviser: Biao Zhang.

Thesis (Ph.D.)--The University of Toledo, 2003.

The logistic regression model is the "back bone" of case-control analysis (Breslow, 1980). Prentice and Pyke (1979) proved that the MLE of log odds ratio parameters is invariant under both prospective and retrospective study designs if the logistic regression model is true. A natural question is how to test the adequacy of the logistic regression model based on retrospective data. Qin and Zhang (1997) and Zhang (1999, 2001) proposed three types of tests. In this dissertation we extend Qin and Zhang's work to the Continuation Ratio Logistic Regression (CRLR) model, one of the commonly used ordinal logistic models with wide applications in the areas of biological, medical and social sciences. We will show that, under the case-control sampling plan, the CRLR model is equivalent to an I-sample selection biased semi-parametric model. We propose empirical likelihood based semi-parametric profile likelihood estimators for both parametric and non-parametric parts. A Kolmogrov-Smirnov type of goodness-of-fit test was proposed to test the adequacy of the CRLR, model based on case-control data. We also present, the asymptotic properties of the estimators. A small simulation study is conducted to assess the power of the proposed test by using a bootstrap procedure. As alternatives, we also proposed two generalized moments based goodness-of-fit, tests for the CRLR model. We also report results on two real data sets along with the bootstrap resampling procedure.

ISBN: 9780496454488Subjects--Topical Terms:

517247
Statistics.

Fitting and testing continuation ratio logistic regression models based on case-control data.
LDR:02341nam 2200265 a 45 001 963255
005 20110831
008 110831s2003 ||||||||||||||||| ||eng d
020 $a 9780496454488
035 $a (UMI)AAI3097835
035 $a AAI3097835
040 $a UMI $c UMI
100 1 $a Peng, Cheng. $3 1286318
245 1 0 $a Fitting and testing continuation ratio logistic regression models based on case-control data.
300 $a 116 p.
500 $a Adviser: Biao Zhang.
500 $a Source: Dissertation Abstracts International, Volume: 64-07, Section: B, page: 3355.
502 $a Thesis (Ph.D.)--The University of Toledo, 2003.
520 $a The logistic regression model is the "back bone" of case-control analysis (Breslow, 1980). Prentice and Pyke (1979) proved that the MLE of log odds ratio parameters is invariant under both prospective and retrospective study designs if the logistic regression model is true. A natural question is how to test the adequacy of the logistic regression model based on retrospective data. Qin and Zhang (1997) and Zhang (1999, 2001) proposed three types of tests. In this dissertation we extend Qin and Zhang's work to the Continuation Ratio Logistic Regression (CRLR) model, one of the commonly used ordinal logistic models with wide applications in the areas of biological, medical and social sciences. We will show that, under the case-control sampling plan, the CRLR model is equivalent to an I-sample selection biased semi-parametric model. We propose empirical likelihood based semi-parametric profile likelihood estimators for both parametric and non-parametric parts. A Kolmogrov-Smirnov type of goodness-of-fit test was proposed to test the adequacy of the CRLR, model based on case-control data. We also present, the asymptotic properties of the estimators. A small simulation study is conducted to assess the power of the proposed test by using a bootstrap procedure. As alternatives, we also proposed two generalized moments based goodness-of-fit, tests for the CRLR model. We also report results on two real data sets along with the bootstrap resampling procedure.
590 $a School code: 0232.
650 4 $a Statistics. $3 517247
690 $a 0463
710 2 $a The University of Toledo. $3 1023352
773 0 $t Dissertation Abstracts International $g 64-07B.
790 $a 0232
790 1 0 $a Zhang, Biao, $e advisor
791 $a Ph.D.
792 $a 2003
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3097835