|
Activity Number:
|
254
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Education
|
| Abstract - #309670 |
|
Title:
|
A SAS Macro Implementation of a Multiple Comparison Post Hoc Test
|
|
Author(s):
|
Alan Elliott*+ and Linda S. Hynan
|
|
Companies:
|
The University of Texas Southwestern Medical Center at Dallas and The University of Texas Southwestern Medical Center at Dallas
|
|
Address:
|
Division of Biostatistics, Dallas, TX, 75390,
|
|
Keywords:
|
Kruskal-Wallis ; nonparametric ; SAS ; post hoc ; multiple comparisons ; Dunn
|
|
Abstract:
|
The Kruskal-Wallis non-parametric analysis of variance is often used instead of a standard ANOVA when samples come from a suspected non-normal population or when variances are non-heterogeneous. This omnibus procedure tests for differences between groups, but provides no specific post hoc pairwise comparisons. This paper provides a SAS macro implementation of a Tukey-type multiple comparison test (Dunn's Test) based on Kruskal-Wallis results from the SAS NPAR1WAY procedure. The implementation can be utilized for up to 10 groups at the 0.05 alpha significance level. In addition, the effectiveness of the procedure to meet a selected type I significance level is evaluated using Monte Carlo simulations.
|
