A comparative study of leftover least squares (LS Residuals) Residuals and sequential (Recursive Residuals And residuum (BLUS Residuals) In testing the problem of heterogeneity and autocorrelation In linear regression models | ||
| Journal of Administration and Economics | ||
| Article 1, Volume 0, Issue 106, October 2018, Pages 271-283 | ||
| Authors | ||
| Prof. Dr. Salma fixed Zakir; Adel Abdul Ghafoor | ||
| Abstract | ||
| This Research has discussed a comparative study of the least square residuals (LS), recursive residuals and the best linear unbiased scalar residuals (BLUS). Tests to the problems of Herero's kedasticity and Autocorrelation have been used .in the use of a regression model represents the relationship between the income and the consumption (k=2) of parameters. A set of programs has been written in Matlab language to find the BLUS residuals . The statistical package (Eviews7) has been used to find recursive residuals. [ It has been found that the least square residuals (LS) have agreed with the recursive residuals and the BLUS residuals in test of Heteroskedosticity, but it has disagreed with them in the Autocorrelation test. | ||
| Keywords | ||
| Statistical axis | ||
|
Statistics Article View: 66 PDF Download: 30 |
||