Estimation of the Parameters of a Linear Functional Relation

SUMMARY A variety of consistent estimators of the parameters of a linear functional relation have been proposed over the years. Some find their basis in least-squares procedures, some in the methods of analysis of variance, some in the method of grouping, and some in diverse other methods. In this paper we generalize a class of estimators suggested originally by the method of grouping, and consider the asymptotic variance of a number of different types of estimators. As a result we are able to make specific recommendations regarding a choice among them.