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

Comparison of two SAS procedures for...

Bursac, Zoran.

Linked to FindBook

Google Book

Amazon

博客來

Comparison of two SAS procedures for longitudinal data with evaluation of SAS experimental procedure MI (multiple imputations).

Record Type:	Electronic resources : Monograph/item
Title/Author:	Comparison of two SAS procedures for longitudinal data with evaluation of SAS experimental procedure MI (multiple imputations)./
Author:	Bursac, Zoran.
Description:	160 p.
Notes:	Source: Dissertation Abstracts International, Volume: 64-03, Section: B, page: 1031.
Contained By:	Dissertation Abstracts International64-03B.
Subject:	Biology, Biostatistics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3083746

Comparison of two SAS procedures for longitudinal data with evaluation of SAS experimental procedure MI (multiple imputations).
Bursac, Zoran.

Comparison of two SAS procedures for longitudinal data with evaluation of SAS experimental procedure MI (multiple imputations). - 160 p.

Source: Dissertation Abstracts International, Volume: 64-03, Section: B, page: 1031.

Thesis (Ph.D.)--The University of Oklahoma Health Sciences Center, 2003.

In this simulation study we compared the performance of SAS procedures MIXED and GENMOD in fitting multiple-population models to longitudinal growth data with missing values. Comparison was done under 40 simulated conditions that included two different percentages of missing data, high and medium correlation structures, four and seven repeated time points and 5 levels of data completeness. Completeness levels included complete data set, data missing at random (MAR), MAR with imputed values, non-ignorable missing data mechanism (NI) and NI with imputed values. Data imputation was performed using SAS experimental procedure MI. We found that PROC MIXED performed with higher accuracy than PROC GENMOD under more than 95% of experimental conditions and that multiple data imputation helped improve model components only for NI missing data mechanism.Subjects--Topical Terms:

1018416
Biology, Biostatistics.

Comparison of two SAS procedures for longitudinal data with evaluation of SAS experimental procedure MI (multiple imputations).
LDR:01771nmm 2200253 4500 001 1858948
005 20041020095007.5
008 130614s2003 eng d
035 $a (UnM)AAI3083746
035 $a AAI3083746
040 $a UnM $c UnM
100 1 $a Bursac, Zoran. $3 1946615
245 1 0 $a Comparison of two SAS procedures for longitudinal data with evaluation of SAS experimental procedure MI (multiple imputations).
300 $a 160 p.
500 $a Source: Dissertation Abstracts International, Volume: 64-03, Section: B, page: 1031.
500 $a Major Professor: Barbara Neas.
502 $a Thesis (Ph.D.)--The University of Oklahoma Health Sciences Center, 2003.
520 $a In this simulation study we compared the performance of SAS procedures MIXED and GENMOD in fitting multiple-population models to longitudinal growth data with missing values. Comparison was done under 40 simulated conditions that included two different percentages of missing data, high and medium correlation structures, four and seven repeated time points and 5 levels of data completeness. Completeness levels included complete data set, data missing at random (MAR), MAR with imputed values, non-ignorable missing data mechanism (NI) and NI with imputed values. Data imputation was performed using SAS experimental procedure MI. We found that PROC MIXED performed with higher accuracy than PROC GENMOD under more than 95% of experimental conditions and that multiple data imputation helped improve model components only for NI missing data mechanism.
590 $a School code: 0361.
650 4 $a Biology, Biostatistics. $3 1018416
690 $a 0308
710 2 0 $a The University of Oklahoma Health Sciences Center. $3 1023966
773 0 $t Dissertation Abstracts International $g 64-03B.
790 1 0 $a Neas, Barbara, $e advisor
790 $a 0361
791 $a Ph.D.
792 $a 2003
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3083746