Activity Number:
|
613
- Practical Applications of Small Area Estimation
|
Type:
|
Topic Contributed
|
Date/Time:
|
Thursday, August 3, 2017 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Survey Research Methods Section
|
Abstract #324622
|
|
Title:
|
Measurement Error in Small Area Estimation: Functional vs. Structural vs. Naive Models
|
Author(s):
|
William Bell* and Gauri Datta and Carolina Franco and Hee Cheol Chung
|
Companies:
|
U.S. Census Bureau and University of Georgia and US Census Bureau and U.S. Census Bureau and University of Georgia
|
Keywords:
|
sample survey ;
area level model ;
covariate ;
prediction
|
Abstract:
|
Small area estimation using area-level models may benefit from covariates observed with error, such as when using other survey estimates as covariates. Given estimated variances of these errors, one can account for uncertainty in the covariates using measurement error models (e.g., Ybarra and Lohr 2008). Two types of measurement error models have been introduced to deal with such errors. The functional measurement error model assumes that the underlying true values of the covariate are fixed but unknown quantities. The structural measurement error model assumes that these true values follow a model, implying a multivariate model for the covariates and the original dependent variable. We compare and contrast these two models under different underlying assumptions. We also explore the consequences for prediction mean squared errors that result from ignoring measurement error when it is present (naïve model) rather than using a functional or structural measurement error model. Comparisons done both analytically and via simulations yield some surprising results. We illustrate the results using data from the U.S. Census Bureau's Small Area Income and Poverty Estimates Program.
|
Authors who are presenting talks have a * after their name.