Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 197 - SPAAC Poster Competition
Type: Topic Contributed
Date/Time: Monday, August 8, 2022 : 2:00 PM to 3:50 PM
Sponsor: Survey Research Methods Section
Abstract #322522
Title: Empirical Best Prediction of Nonlinear Parameters for an Informative Sample Design
Author(s): Yanghyeon Cho* and Emily Berg
Companies: Iowa State University and Iowa State University
Keywords: Empirical best predictor; Parametric Bootstrap; Bias correction; Nonlinear parameter; Sampling weight; Informative sampling
Abstract:

This material proposes the predictors and their prediction error of nonlinear small-area parameters under an informative probability sampling design for both sampled and nonsampled areas. Our approach uses the relationships between the weights and the population defined in Pfeffermann and Sverochkov (2007) to obtain the Monte Carlo approximation of an empirical best-unbiased predictor (EBUP), as in Molina and Rao (2010). We propose mean-squared error estimators that do not require a double bootstrap procedure to assess the prediction error. We also avoid using a fully parametric bootstrap by generating the plausible model parameter estimates with the GEE method. We illustrate the bias-corrected confidence interval suggested by Carlin and Gelfand (1990) in the context of the small area estimation problems. In the simulation study, we compare the suggested methods with the existing methods, supporting the good performance of the proposed methods. We illustrate an application using the Conservation Effects Assessment Project data.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2022 program