|
Activity Number:
|
573
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 6, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #302956 |
|
Title:
|
Penalized Balanced Sampling: Nonparametric Guidance for Survey Design
|
|
Author(s):
|
Jay Breidt*+ and Guillaume Chauvet
|
|
Companies:
|
Colorado State University and National Institute for Statistics and Economic Studies
|
|
Address:
|
102 Statistics, Fort Collins, CO, 80523-1877,
|
|
Keywords:
|
linear mixed effect models ; penalized splines ; cube algorithm ; nonparametric
|
|
Abstract:
|
Linear mixed effects models (LMEs) encompass many statistical methods and have found extensive use in estimation for complex surveys, particularly in small area estimation and in various extensions of generalized regression estimation (GREG), including nonparametric regression. They have also been used as a means of relaxing constraints in calibration estimation. The purpose of this work is to consider methods by which LMEs may be used at the design stage of a survey. We review the ideas of balanced sampling and suggest an implementation of the cube algorithm for which "penalized balanced samples" can be selected. Such samples have the property that Horvitz-Thompson estimators from penalized balanced samples behave like LME model-assisted regression estimators. Empirical results with "nonparametric" survey designs demonstrate the usefulness of the methodology.
|
