Comparison of regression equations in health data

Abstract

In many situations, the variable answer Y and the set of regression variables are measured in a set composed of different groups. Authors' objective is to examine how groups differ among themselves by their Y and X_i relationship and to determine whether the set of regression curves are parallel, whether they have a common intercept or whether they are identical. Dummy of variables is used to identify groups in data set. Kleinbaum et al. (1998), Krzanowski (1998), Neter, J., et al. (1996) and Seber, G. A. F. (1977) compare several regression equations by models of multiple linear regression. The method supplies the same information as that obtained with covariance and variance analyses. Methodology was applied to data of blood donors at the Blood Bank of the University Hospital of Maringá 1995-96