|
Activity Number:
|
28
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #305748 |
|
Title:
|
Performance of Robust and Nonrobust Roy-Bargmann Stepdown Follow up to a Significant MANOVA under a Variety of Conditions: a Simulation Study
|
|
Author(s):
|
Holmes Finch*+
|
|
Companies:
|
Ball State University
|
|
Address:
|
Educational Psychology, Muncie, IN, 47306,
|
|
Keywords:
|
MANOVA ; post hoc analysis ; Roy-Bargmann ; stepdown procedures
|
|
Abstract:
|
Following a significant MANOVA, researchers may want to identify individual dependent variables for which group means differ. One approach for examining these univariate differences is the Roy-Bargmann stepdown procedure. In this analysis dependent variables are ordered by theoretical importance, after which sequential univariate testing is conducted in descending order of importance, where more important variables serve as covariates for less important ones. This simulation study examined the ability of the Roy-Bargmann method to correctly identify dependent variables for which group means differ (power) and those for which means do not differ (Type I error). Several factors were manipulated in the simulations. Results suggest that the manipulated variables do influence the performance of the method, and that robust and non-robust approaches perform differently under some conditions.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2006 program |
