|
Activity Number:
|
222
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #304130 |
|
Title:
|
Comparison of Prediction Interval Methods for a One-Way Random Effects Model
|
|
Author(s):
|
Jamie M. Baldwin*+ and Ramon Littell
|
|
Companies:
|
Info Tech, Inc. and University of Florida
|
|
Address:
|
5700 SW 34th Ave, Gainesville, FL, 32608,
|
|
Keywords:
|
linear mixed models ; EBLUPs ; prediction interval
|
|
Abstract:
|
Empirical Best Linear Unbiased Predictors (EBLUPs) for analyzing Linear Mixed Models (LMMs) are widely used, yet the best way to evaluate the precision of the EBLUPs is not generally understood. We study this issue in the context of formulating a prediction interval for the realized value of a random effect in the balanced one-way random effects model. Four prediction intervals are compared in a Monte Carlo simulation study to determine which method has the most accurate coverage rate. While the structure of the four methods compared is similar, the disparity comes in the impact of the parameterization, and the estimate of the standard error which subsequently impacts the degrees of freedom. While two of the methods are currently available in SAS, the other methods studied often outperform the currently available methods.
|
- 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 2009 program |
