Goodness of Fit in Measurement Error Model

Abstract

Various types of models can be fitted to a given set of data. ?A natural question is how to choose the best fitted model. ?The coefficient of determination, popularly known as "R-square statistic", is used to judge the goodness of fit in a usual multiple linear regression model. ?This R-square statistic depends on the ordinary least squares estimator of regression coefficient and works well only when the data is correctly observed without any measurement error. ?When the data is contaminated with measurement errors, then the R-square statistic does not work and gives incorrect results about the goodness of fit about the model. ?How to judge the goodness of fit in the measurement error models is the question is answered in this talk.?