|
Activity Number:
|
70
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #308890 |
|
Title:
|
A Simple Computational Method for Estimating Mean Squared Prediction Error in General Small-Area Model
|
|
Author(s):
|
Snigdhansu Chatterjee*+ and Partha Lahiri
|
|
Companies:
|
The University of Minnesota and University of Maryland
|
|
Address:
|
School of Statistics, Minneapolis, MN, 55455,
|
|
Keywords:
|
Small area ; mean squared prediction error ; bootstrap ; GLMM
|
|
Abstract:
|
A general small area model is a hierarchal two stage model, of which special cases are mixed linear models, generalized linear mixed models and hierarchal generalized linear models. In such models, the variability of predictors (like the empirical best predictor or the empirical best linear unbiased predictor) is usually quantified with their mean squared prediction error (MSPE). Estimators for MSPE are generally not available outside some special cases. We propose a simple resampling based estimation of MSPE for any general small area model. The proposed MSPE estimator has high order accuracy and can be guaranteed to be positive.
|
