A class of estimators of the parameters in linear regression with censored data (Q1118946)

This paper considers the linear regression model \(y_ i=\alpha +\beta x_ i+\epsilon_ i\) \((n=1,...,n)\) where the \(y_ i's\) may be censored and therefore not completely observable. A new class of estimators is proposed (it includes several estimators previously considered in the literature). The paper gives sufficient conditions under which the proposed estimators are strongly consistent and asymptotically normal.