Weak \(\sqrt n\)-consistency of the least weighted squares under heteroscedasticity. (English) Zbl 1228.62026

Summary: Weak \(\sqrt{n}\)-consistency of the least weighted squares estimator of the coefficients of a regression model is proved generally under the heteroscedasticity of error terms. The assumptions required for the weak \(\sqrt{n}\)-consistency are briefly discussed. The roots of the heteroscedasticity are also critically considered.


62F12 Asymptotic properties of parametric estimators
62J05 Linear regression; mixed models
62H12 Estimation in multivariate analysis