Estimasi Model Regresi Panel Komponen Error Satu Arah dengan Metode Maksimum Likelihood
(1) Fakultas Sains dan Teknologi, Universitas Airlangga
(*) Corresponding Author
One way error component panel regression model is regression model using panel data that has double index in its variable to test unobservation individual effect by taking individual randomly from one population. The purpose of this final project is to estimated one way error component panel regression model with maximum likelihood method. The result of parameter regression estimation still depends on and components so that, the iteration process until obtained the convergent parameter vector is needed to estimated the parameter. Fisher Test is used for testing the absence of individual effect. Likelihood Ratio Test is used for testing the influence of all of all predictor variables concurrently and each individually.The data for the application of the one way error component panel regression model with maximum likelihood method are human development index in some districs / cities of East Java Province for 2007-2010 as the dependent variable ( ), while predictors variable include: the percentage of opened unemployment rate ( ), the percentage of baby mortality ( ), the percentage of illiteracy for over 15 years old ages. The lack of fit test of one way error component panel regression model for α = 0.05 shows that the model is appropriate and significant predictor variable are with , and error random
Baltagi, B. H. (2005), “Econometric analysis of panel data”, 3th edition, TechBooks, New Delhi, India.
BPS. 2010. Indikator Makro Ekonomi dan Sosial Provinsi Jawa Timur. Jakarta : BPS.
Gujarati, D. (2004), “Basic Econometrics”, 4th edition, McGraw-Hill, New York.
Hsiao, C. (2003), “Analysis of Data Panel”, 2nd edition, Cambridge University Press, West Nyack, NY, USA.
Matyas, L. and Sevestre, P., 2008, “The Econometrics of Panel Data”, Springer-Verlag Berlin Heidelberg.
Article MetricsAbstract view : 0 times
PDF - 0 times
- There are currently no refbacks.