The effect of two different calculation methods for obtaining relative impact indicators is modelled. Science policy considerations make it clear that evaluating the sets of publications, the “ratio of the sums” method should be preferred over the “mean of the ratios” method....