Random regression coefficient area level models
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21340%2F12%3A00199080" target="_blank" >RIV/68407700:21340/12:00199080 - isvavai.cz</a>
Výsledek na webu
—
DOI - Digital Object Identifier
—
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Random regression coefficient area level models
Popis výsledku v původním jazyce
The Fay-Herriot (FH) model is a basic area level model. It is a special instance of the linear mixed models with fixed and random effects. In this contribution we introduce the generalized FH model which, unlike the classical FH model, includes random regression coefficients to treat situations where small areas are divided into several groups, and where direct estimators of the variable of interest follow different relation depending on group. We use the REML (Restricted Maximum Likelihood) method to obtain estimates of model parameters. We provide formulas to calculate EBLUP (Empirical Best Linear Unbiased Predictor) of the variable of interest and to estimate its mean squared error. Simulation experiments are presented to investigate the behaviour of the REML estimates and to~show the accuracy of the EBLUPs calculated by the proposed model. Finally, the FH model and its proposed generalization are compared. The FH model proves to be very adaptable to the data generated by the model
Název v anglickém jazyce
Random regression coefficient area level models
Popis výsledku anglicky
The Fay-Herriot (FH) model is a basic area level model. It is a special instance of the linear mixed models with fixed and random effects. In this contribution we introduce the generalized FH model which, unlike the classical FH model, includes random regression coefficients to treat situations where small areas are divided into several groups, and where direct estimators of the variable of interest follow different relation depending on group. We use the REML (Restricted Maximum Likelihood) method to obtain estimates of model parameters. We provide formulas to calculate EBLUP (Empirical Best Linear Unbiased Predictor) of the variable of interest and to estimate its mean squared error. Simulation experiments are presented to investigate the behaviour of the REML estimates and to~show the accuracy of the EBLUPs calculated by the proposed model. Finally, the FH model and its proposed generalization are compared. The FH model proves to be very adaptable to the data generated by the model
Klasifikace
Druh
O - Ostatní výsledky
CEP obor
BB - Aplikovaná statistika, operační výzkum
OECD FORD obor
—
Návaznosti výsledku
Projekt
—
Návaznosti
S - Specificky vyzkum na vysokych skolach
Ostatní
Rok uplatnění
2012
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů