|
Activity Number:
|
46
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 11, 2002 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
General Methodology
|
| Abstract - #301751 |
|
Title:
|
The Application of Bootstrap Methodology to Multilevel Mixed Effects Linear Models under Conditions of Error Term Nonnormality
|
|
Author(s):
|
Yann-Yann Shieh*+ and Rachel Fouladi+
|
|
Affiliation(s):
|
American Institute of Research and U. T. M. D. Anderson Cancer Center
|
|
Address:
|
1990 K St. NW, Suite 500, Washington, District of Columbia, 20006, 1515 Holcombe Blvd, Box 243, Houston, Texas, 77030, U.S.A.
|
|
Keywords:
|
multilevel modeling ; bootstrap
|
|
Abstract:
|
Current statistical practice by the general research community when analyzing multilevel data is largely suboptimal. Recent examinations of the performance characteristics of multilevel statistics has shown the non-robustness of some of the model parameter estimates (Shieh, Fouladi, & Pullum, 1999). Bagaka (1989) provided a first step in understanding the impact of bootstrap in multi-level data analysis contexts; however, a more thorough examination is necessary. This paper represents a second step in that direction. The purpose of this study is to use a bootstrap resampling approach as it is applied to multilevel mixed effects linear models to get improved estimates and test statistics. The bootstrap provides an alternative set of procedures under almost no distributional assumptions. In this paper, we examine the application of bootstrap methodology to multilevel models under conditions which normality assumptions are violated.
|
- 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 2002 program |
