mspeFHlin {saeMSPE} | R Documentation |
Compute MSPE through linearization method for Fay Herriot model
Description
This function returns MSPE estimator with linearization method for Fay Herriot model. These include the seminal Prasad-Rao method and its generalizations by Datta-Lahiri, Datta-Rao-Smith and Liu et.al. All these methods are developed for general linear mixed effects models.
Usage
mspeFHlin(Y, X, D, method = "PR", var.method = "default")
mspeFHPR(Y, X, D, var.method = "default")
mspeFHDL(Y, X, D, var.method = "default")
mspeFHDRS(Y, X, D, var.method = "default")
mspeFHMPR(Y, X, D, var.method = "default")
Arguments
Y |
(vector). It represents the response value for Fay Herriot model. |
X |
(matrix). It stands for the available auxiliary values. |
D |
(vector). Stands for the known sampling variances of each small area levels. |
method |
The MSPE estimation method to be used. See "Details". |
var.method |
The variance component estimation method to be used. See "Details". |
Details
Default method
for mspeFHlin
is "PR" ,proposed by N. G. N. Prasad and J. N. K. Rao, Prasad-Rao (PR) method uses Taylor series expansion to obtain a second-order approximation to the MSPE. Function mspeFHlin
also provide the following methods:
Method "DL" proposed by Datta and Lahiri , It advanced PR method to cover the cases when the variance components are estimated by ML and REML estimator. Set method = "DL"
.
Method "DRS" proposed by Datta and Smith, It focus on the second order unbiasedness appoximation when the variance component is replaced by Empirical Bayes estimator. Set method = "DRS"
.
Method "MPR" is a modified version of "PR", It was proposed by Liu et al. It is a robust method that broaden the mean function from the linear form. Set method = "MPR"
.
Default var.method
and available variance component estimation method for each method is list as follows:
For method = "PR"
, var.method = "MOM"
is the only available variance component estimation method,
For method = "DL"
, var.method = "ML"
or var.method = "REML"
is available,
For method = "DRS"
, var.method = "EB"
is the only available variance component estimation method,
For method = "MPR"
, var.method = "OBP"
is the only available variance component estimation method.
Value
This function returns a list with components:
MSPE |
(vector) MSPE estimates for Fay Herriot model. |
bhat |
(vector) Estimates of the unknown regression coefficients. |
Ahat |
(numeric) Estimates of the variance component. |
Author(s)
Peiwen Xiao, Xiaohui Liu, Yuzi Liu, Jiming Jiang, and Shaochu Liu
References
N. G. N. Prasad and J. N. K. Rao. The estimation of the mean squared error of small-area estimators. Journal of the American Statistical Association, 85(409):163-171, 1990.
G. S. Datta and P. Lahiri. A unified measure of uncertainty of estimated best linear unbiased predictors in small area estimation problems. Statistica Sinica, 10(2):613-627, 2000.
G. S. Datta and R. D. D. Smith. On measuring the variability of small area estimators under a basic area level model. Biometrika, 92(1):183-196, 2005.
X. Liu, H. Ma, and J. Jiang. That prasad-rao is robust: Estimation of mean squared prediction error of observed best predictor under potential model misspecification. Statistica Sinica, 2020.
Examples
X = matrix(runif(10 * 3), 10, 3)
X[,1] = rep(1, 10)
D = (1:10) / 10 + 0.5
Y = X %*% c(0.5,1,1.5) + rnorm(10, 0, sqrt(2)) + rnorm(10, 0, sqrt(D))
mspeFHlin(Y,X,D,method = "PR", var.method = "default")