ESTIMATION OF THE NUMBER OF UNEMPLOYED AND EMPLOYED FOR LITHUANIAN COUNTIES AND MUNICIPALITIES USING MODELS

Linas Naujanis, Danutė Krapavickaitė

Abstract


Problems of finite population parameters estimation are analyzed in this paper. Four methods have been used for parameterestimation: sampling design-based unbiased estimator, multiple regression and logistic regression model-based estimators and James–Stein estimator. The design-based estimator is unbiased, but its standard deviation is usually high. Model-based estimators are notunbiased, but their standard deviations are low. In order to minimize the standard deviation and the bias, the James–Stein estimator isapplied. Labour force survey data of Statistics Lithuania are used for simulation to study model-based estimators for the number ofunemployed and employed persons in districts and counties, and the role of information on registered unemployment in these models.

Keywords


area with small sample size, model-based estimator, multiple regression model, James–Stein estimator.

Full Text: PDF

Lithuanian Journal of Statistics ISSN 1392-642X, eISSN 2029-7262
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 License.