|
Activity Number:
|
28
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #306540 |
|
Title:
|
Generation of the Distribution of the Test for a Latin Square Design with Heterogeneous Variances
|
|
Author(s):
|
Miin-Jye Wen*+ and Hubert Chen
|
|
Companies:
|
National Cheng Kung University and National Cheng Kung University
|
|
Address:
|
Management Building, 2nd Floor, No. 1 Ta-Hsueh Road, 70101 Tainan, , Taiwan
|
|
Keywords:
|
heteroscedasticity ; one-stage procedure ; t distribution ; Latin square ; simulation ; critical values
|
|
Abstract:
|
Assuming a Latin square design model whose error terms follow independent and normal distributions with unknown and possibly unequal variances, the interest is to develop a one-stage sampling procedure to test the null hypothesis of equal treatment effects. A weighted sample mean is used to serve as a basis for a test statistic for the Latin square design model. It is found that the distribution of the test statistic based on the weighted sample mean is completely independent of the unknown variances. A SAS simulation program to m by m Latin square design is developed to implement the testing procedure.
|