Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Statistics and Econometrics

Panel data can be considered as one of the most important topics nowadays. As it includes both dimensions of time and cross section. And that is in contrast to time series and cross sectional data sets. The time dimension in panel data is represented in the observations collected over time periods on each individual. For forecasting purposes the time series should be stationary. Unit roots problem hinders reduces the efficiency of any forecasting process. There are two main approaches developed to flag the panels for having unit roots. The first approach depends on performing a certain test on the panels, then collecting the information to get an overall idea on the existence of unit roots in panel data, such as: Augmented Dickey Fuller (ADF) test, Im, persran and shin (IPS) test, and combing P-value test. The second approach detects the unit roots effect by running a certain test once on all the panels, such as: Levin, lin and Chu (LLC) test and breitung test

Issued also as CD