Master'sOpen Access

Seemingly unreleated regression equations and an application to export function of Turkey

1998
0 views
0 downloads
Advisor: Prof. Dr. Bedriye Saraçoğlu

Abstract (EN)

In this study; Seemingly Unrelated Regression Equations(SURE) which forms the error terms related equation system is examined in order to show its efficient over the Ordinary Least Square Method (OLS). Towards this aim first of all a brief introduction about OLS is given and the Generalized Least Square method that constitutes the main part of SURE is also examined. After that SURE method's definition, spesifications and substraction of its estimator are shown and gain in the efficiency related to OLS method is proved. In empirical analysis; Turkey's exports to 18 OECD countries between 1982-1995 was estimated by using OLS and SURE methods and found out that the estimators of SURE method were more efficient. Conclusively, by applying SURE method to the models; which seems to be independent but in which there exists a relation between its error terms of equation, it has been proved that empirical and teoric results of SURE estimator is more efficient than OLS estimator.

Author

Nükhet Doğan

How to Cite

Nükhet Doğan (Master Thesis). Seemingly unreleated regression equations and an application to export function of Turkey, 1998, Gazi University.

License

Tüm Hakları Saklıdır

This work is shared under the specified license terms.

More theses from Gazi University