|
Activity Number:
|
509
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #306808 |
|
Title:
|
New Methods Using Levene-Type Tests for Hypotheses about Dispersion Differences
|
|
Author(s):
|
Xiaoni Liu*+ and Dennis A. Boos and Cavell Brownie
|
|
Companies:
|
North Carolina State University and North Carolina State University and North Carolina State University
|
|
Address:
|
Department of Statistics, 601 Jones Ferry Road, Carrboro, NC, 27510,
|
|
Keywords:
|
Levene type test ; RCB designs ; bootstrap ; robustness ; dispersion differences ; Huber m-estimation
|
|
Abstract:
|
Levene type tests are well known to be robust tests for equality of scale in one-way designs. Currently, the available Levene type tests for two-way designs with one observation per cell (i.e., RCB designs) employ either standard ANOVA F tests on the absolute values of least squares (OLS) residuals, or weighted least squares (WLS) ANOVA F tests on the OLS residuals. In this paper, we develop new methods based on Levene type tests for two-way RCB designs that can assess equality of variances across blocks and/or treatments. We use OLS ANOVA and WLS ANOVA F tests on the absolute values of residuals obtained from models fit by Huber M-estimation. We also apply bootstrap methods to these Levene type tests and compare these Levene type tests in terms of robustness and power by simulation. In result, the new methods compare favorably with older methods.
|
