|
A Specialized Empirical Bayes Single Regression Coefficient Estimation Procedure for Panel Data Linear Regression Models |
|---|---|
| รหัสดีโอไอ | |
| Creator | Khan Bahadar, Eatzaz Ahmed, Shahid Akbar |
| Title | A Specialized Empirical Bayes Single Regression Coefficient Estimation Procedure for Panel Data Linear Regression Models |
| Contributor | - |
| Publisher | TuEngr Group |
| Publication Year | 2565 |
| Journal Title | International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies |
| Journal Vol. | 13 |
| Journal No. | 4 |
| Page no. | 13A4G: 1-12 |
| Keyword | Empirical Bayes, Prior Mean, Prior Variance, Zellner's g-priors, Prior Precision, Keynesian consumption function. |
| URL Website | http://TuEngr.com/Vol13-4.html |
| Website title | ITJEMAST V13(4) 2022 @ TuEngr.com |
| ISSN | 2228-9860 |
| Abstract | In this article, a specialized empirical Bayes estimator for the panel data model is applied to real-world data. This estimator is suitable for all the standard classical panel data models. In this study, the simple linear regression model with orthogonal regressors has been used, and the coefficients for the ith unit of the panel have been estimated individually and independently, that is, the intercepts and the slopes coefficients for each unit of the panel are estimated as scalar quantities rather than a vector of both. The estimator used here is, "the precision weighted arithmetic mean of the ordinary least squares coefficient estimates of the kth regressor across all the units of the panel as the estimate of the prior mean and the Zellner's g-prior is used as the estimate of the prior variance and the resultant prior precision". By taking different values by the prior precision parameter, it has the potential to produce all the standard frequentist panel data linear regression model estimators. The estimate of the prior precision parameter is obtained from the data corresponding to all units of the panel therefore, it is considered to be the most reliable. |
