A critical evaluation of different comparator methods is given. The method based on experimentally-determined and convertible comparison factors (k) is considered to be the most suitable for general use. An alternative method is proposed, introducing generalized k0 factors which are independent of irradiation and measuring conditions. This approach combines the simplicity of the absolute methods with nearly the same accuracy attained by the relative ones. It is suggested that k0 factors be compiled in all cases when using single-comparator methods, to allow a continuous re-evaluation.