|
Activity Number:
|
428
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #307492 |
|
Title:
|
Resampling Methods in Asymptotic Statistical Inferences
|
|
Author(s):
|
Fassil Nebebe*+ and Tak K. Mak
|
|
Companies:
|
Concordia University and Concordia University
|
|
Address:
|
1455 de Maisonneuve Blvd., Montreal, PQ, H3G1M8, Canada
|
|
Keywords:
|
resampling designs ; intraclass correlation ; heteroscedasticity
|
|
Abstract:
|
The bootstrap is known to be useful in producing asymptotic variances of estimators which are otherwise difficult to obtain analytically. This application can be made more efficient by special resampling designs based on linear regression models. The present paper studies the use of these designs to derive efficiently the sampling distribution of any statistic for any sample size. The theory developed will be applied to some well known statistical inference problems, such as intraclass correlation estimation and the test for.
|
