Regular estimators for subclasses of linear estimators (Q1088318): Difference between revisions

The concept of regular estimator is due to \textit{J. Roy} and \textit{I. M. Chakravarti} [Ann. Math. Stat. 31, 392-398 (1960; Zbl 0093.158)]. For simple random sampling with or without replacement, regular estimators exist in the \(T_ 1\)-class of linear estimators. However, for sampling with varying probabilities, a regular estimator does not exist in the \(T_ 1\), \(T_ 2\) and \(T_ 3\) classes of linear estimators. A regular estimator exists in the \(T_ 4\) class of linear estimators, if the sampling scheme is ''balanced''.