ARTICLE
ABSTRACT
The main objective of this paper is to demonstrate the usefulness of two-level modeling methodology for estimating the socio-economic variables. In the first part of the paper one will present the model construction stages taking the two-level structure of population and variables into account. In the second part an example of using the above methodology for estimating the number of working people in crosssection of counties is presented. Province were chosen as second-level unit.
With the two-level modeling methodology one can include variation of the level of the considered variable and the strength of its dependencies with explanatory variables between the groups. Furthermore, additional information has been obtained by using the explanatory variables from the second level – concerning the entire group.
In the second part, as the explanatory variables, among others, the results of unique study in the Statistical Office in Poznan which concerned the flow of employees were used. The main source of information of this study are the fiscal records of the Ministry of Finance. These data concern the year 2006 and this has been the first information for commuting since 1988 provided by the Central Statistical Office.
The comparison of the quality of estimates obtained using two-level approach and the classical linear regression were conducted. The results show the advantage of two-level model.
KEYWORDS
two-level modeling, ANOVA, random component, commuting
REFERENCES
[1] Bliese P., (2012), Multilevel Modeling in R (2.4) A Brief Introduction to R, the multilevel package and the nlme package, Paul Bliese, April 10, http://cran.r-project.org/doc/contrib/Bliese Multilevel.pdf.
[2] Bołt T., Krauze K., Kulawczuk T., (1985), Agregacja modeli ekonometrycznych, Panstwowe Wydawnictwo Ekonomiczne, Warszawa.
[3] Goldstein H., (2003), Multilevel Statistical Models, 3rd edition, London: Edward Arnold.
[4] Harville D.A., (1974), Bayesian Inference for Variance Components Using Only Error Contrasts, Biometrika, 61, 383-385,
[5] Hox J., (2002), Multilevel Analysis. Techniques and Applications, Lawrence Erlbaum Associates, Publishers, London.
[6] Klimanek T., (2003), Wielopoziomowa analiza struktury agrarnej gminy w systemie Geo-Info, Praca doktorska napisana na Akademii Ekonomicznej w Poznaniu na Wydziale Zarzadzania w Katedrze Statystyki, Poznan.
[7] Kopczewska K., Kopczewski T., Wójcik P., (2009), Metody ilosciowe w R Aplikacje ekonomiczne i finansowe, CeDeWu.pl, Warszawa.
[8] Krzysko M., (1996), Statystyka matematyczna, Wydawnictwo Naukowe UAM, Poznan.
[9] Lin X., (1997), Variance Component Testing in Generalized Linear Models with Random Effects, Biometrika, 84, 309-25.
[10] Pinheiro J.C., Bates D.M., (2000), Mixed-Effects Models in S and S-PLUS, New York: Springer-Verlag.
[11] Rao J.N.K., (2003), Small Area Estimation, Wiley & Sons, New York.
[12] Raudenbush S.W., Bryk A.S., (2002), Hierarchical Linear Models. Applications and Data Analysis Methods, Second Edition, Sage Publications, London Thousand Oaks New Delhi.
[13] Rydlewski J.P., (2009), Estymatory najwiekszej wiarygodnosci w uogólnionych modelach regresji nieliniowej, Praca doktorska napisana na Uniwersytecie Jagiellonskim na Wydziale Matematyki i Informatyki w Instytucie Matematyki, Kraków.
[14] Sakamoto Y., Ishiguro, M., and Kitagawa G., (1986), Akaike Information Criterion Statistics, D. Reidel Publishing Company.
[15] Schwarz G., (1978), Estimating the Dimension of a Model, Annals of Statistics, 6, 461-464.
[16] Twisk J.W.R., (2010), Analiza wielopoziomowa – przykłady zastosowan Praktyczny podrecznik biostatyki i epidemiologii, Oficyna Wydawnicza SGH, Warszawa.
[17] Weziak D., (2007), Wielopoziomowe modelowanie regresyjne w analizie danych, Wiadomosci Statystyczne, 9 (556), 1-12.
