TWO JACKKNIFE ESTIMATORS FOR THE SIGNAL AND COMPARISON OF THEIR PERFORMANCE BY A MONTE CARLO SIMULATION

Abstract

The paper introduces two jackknife estimators of the signal. The mean square errors of the proposed estimators and two other estimators of the signal are studied by simulation. The results of the study show that in designs baving weak to mild collinearities both jackknife estimators are more efficient than the naive estimator as well as non-negative minimum biased estimator. The non-negative minimum biased estimator performs better under strong collinearity.