Abstract
For the analyses of dose-relations of radiation-induced chromosome aberrations a weighted least squares method has to be carried out since the variance of observed aberration yields are different at different doses. Consequently identical statistical weights have to be used for the comparison of dose-response functions. For this reason a weighted identity test is presented. The derivation of the test quantitity is described in a generalized form. The practical application of the test and the computation of the test quantitity is shown for the linear and linear-quadratic model.