الفهرس 
Estimation of the unbalanced panel data models
ContentOnly 14 pages are availabe for public view
 |  | from | 146 |  |  |
| | | from | 146 | | |
Abstract
Panel data set can be considered as one of most important data sets because it allows the researcher to look at dynamic relationships and that would not be happened with a single cross section and time series. Now there is a big problem appeared which is missing values in panels and that is lead us to unbalanced panel data models. Our thesis is interested to deal with this kind of data in case of unbalanced and balanced with different assumptions like correlated and Non correlated effects. In panel data analysis there are two approaches fixed effect and random effect models. Fixed Effect (FE) model is useful when there is correlation between individual effect and explanatory variable. Random Effect (RE) model is useful when there is no correlation between individual effect and explanatory variable. Also the methods of estimation for missing data affected mean square error (MSE) and R-Squared, so we used three methods to estimate the missing values and comparing the results. The main goal of the thesis is to compare between the results of balanced and Unbalanced panel data under different two assumptions correlated and Non-correlated effects