Summary/Abstract |
The purpose of this note is to point out that some of the well-known pseudo-generalized least squares estimators (e.g., the two-step and interactive Cochrane-Orcutt Estimators (1949), the two-step and interactive Prais-Winsten Estimators (1954), The Durbin Estimator (1960) etc.) are unbiased under very reasonable conditions,provided their means exist.
|