FHme: Small area estimation when the covariates subject to sampling...

In aminfathullah/saeme: Small Area Estimation with Measurement Error

Description

- Calculate the estimated value of Y with auxiliary variable X where the auxiliary variable X also measured with error
- Calculate the MSE of estimated Y with jackknife method

Usage

    FHme(y, x, vardir, C, iter)

Arguments

y	vector of direct estimation for each area
x	matrix of auxiliary variable from another survey. It contains two or more columns. first column is vector of 1 (intercept). It should be like cbind(1, aux_var). See example.
vardir	vector of design variance of variable y
C	matrix of variance of variable x. It has same column number like x. The first column is vector of 0 (intercept variance). It should be like cbind(0, variance_x). See example
iter	Number of iteration for finding w. The dafault is 2. See reference for more details.

Value

y_me	small area estimation for each area
gamma	estimation of gamma for each area
sigma	estimation of sigma for each area
beta	estimation of beta coefficient
psi	MSE of direct estimator
mse	MSE of small area estimation
call	function call

Author(s)

Amin Fathullah ([e-mail])

References

Ybarra, L. M., & Lohr, S. L. (2008). Small area estimation when auxiliary information is measured with error. Biometrika, 919-931.

Examples

data(bmi)

#calculate SAE with measurement error
sae_me <- FHme(y = bmi$y, x = cbind(1, bmi$x), vardir = bmi$mse_y, C = cbind(0, bmi$mse_x))

sae_me
summary(sae_me)
plot(sae_me)

aminfathullah/saeme documentation built on July 12, 2019, 7:46 a.m.