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

Optimal analysis of group randomized...

Braun, Thomas Michael.

Linked to FindBook

Google Book

Amazon

博客來

Optimal analysis of group randomized trials with permutation tests.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Optimal analysis of group randomized trials with permutation tests./
Author:	Braun, Thomas Michael.
Description:	107 p.
Notes:	Chair: Ziding Feng.
Contained By:	Dissertation Abstracts International60-08B.
Subject:	Biology, Biostatistics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9944097
ISBN:	0599458208

Optimal analysis of group randomized trials with permutation tests.
Braun, Thomas Michael.

Optimal analysis of group randomized trials with permutation tests. - 107 p.

Chair: Ziding Feng.

Thesis (Ph.D.)--University of Washington, 1999.

Two facts complicate the analysis of group randomized trials (GRTs). First, individual outcomes within each group are often correlated. Second, the number of groups in a GRT is often not sufficient to make asymptotic approximations possible. Therefore, methods such as GEE (Liang & Zeger, 1986) and PQL (Breslow & Clayton, 1993), originally developed for longitudinal studies, may not be valid.

ISBN: 0599458208Subjects--Topical Terms:

1018416
Biology, Biostatistics.

Optimal analysis of group randomized trials with permutation tests.
LDR:02534nam 2200289 a 45 001 931044
005 20110429
008 110429s1999 eng d
020 $a 0599458208
035 $a (UnM)AAI9944097
035 $a AAI9944097
040 $a UnM $c UnM
100 1 $a Braun, Thomas Michael. $3 1254592
245 1 0 $a Optimal analysis of group randomized trials with permutation tests.
300 $a 107 p.
500 $a Chair: Ziding Feng.
500 $a Source: Dissertation Abstracts International, Volume: 60-08, Section: B, page: 3658.
502 $a Thesis (Ph.D.)--University of Washington, 1999.
520 $a Two facts complicate the analysis of group randomized trials (GRTs). First, individual outcomes within each group are often correlated. Second, the number of groups in a GRT is often not sufficient to make asymptotic approximations possible. Therefore, methods such as GEE (Liang & Zeger, 1986) and PQL (Breslow & Clayton, 1993), originally developed for longitudinal studies, may not be valid.
520 $a As an alternative, this thesis develops randomization-based methods for analyzing GRTs. First, we review the general validity of permutation tests and demonstrate their validity when specifically applied to GRTs. Second, we extend earlier work of Lehmann & Stein (1949) and develop a permutation test for correlated data that is most powerful within the class of all permutation tests. The statistic is a weighted sum of cluster residuals, where the weights are a function of the group sizes and the variability of the individual outcomes. Third, we propose an algorithm to compute permutation-based confidence intervals from our recommended test.
520 $a Next, simulations of correlated binary and continuous outcomes demonstrate that: (1) permutation tests have nominal size, unlike GEE, which is too liberal in small GRTs, (2) the power of our test equals that of GEE once the test in GEE is adjusted to be nominal, and (3) permutation-based confidence intervals have excellent coverage probabilities and are not necessarily symmetric. The importance of weighting in permutation tests is also demonstrated. Last, we analyze two recently published GRTs, discuss the findings based on permutation tests, and compare those findings with GEE and PQL.
590 $a School code: 0250.
650 4 $a Biology, Biostatistics. $3 1018416
690 $a 0308
710 2 0 $a University of Washington. $3 545923
773 0 $t Dissertation Abstracts International $g 60-08B.
790 $a 0250
790 1 0 $a Feng, Ziding, $e advisor
791 $a Ph.D.
792 $a 1999
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9944097