An estimator is presented for the coefficients of the quadratic errors-in-variables (EV) model when replicated observations at some experimental points.

This paper studies,in case that replicated observations are available in some experimental points,the semi linear errors in variables model in which independent variables are non random.

Multivariate random coefficients model of repeated measures data in medical research;

Tests for presuppositions concerning validity of split-plot analysis of variance with repeated measures;

A new method is given to choose the best values in the multiple observations, and the higher precision of adjusting result could be taken with this way.

In order to solving the question that the best observations would be influenced by worse ones and the mean calculated by the method wouldn’t be the best, a new method is given to choose the best values in the multiple observations based on the specific properties of the multiple observed values.

In this paper, a linear varying-coefficients structural EV model with replicated observations is given.