The bibliometric measurement process transfers scientific publications and citations into indicators on scientific impact. In defining specific measurement paths researchers hold several degrees of freedom as various methodical decisions are scarcely founded on stringent criteria and their respective implications are not fully understood.
These diverse measurement paths result in varying measurements. We propose to compute many possible measurement paths and to analyse the resulting measurement variation. On the one hand, effects of decisions can thus be better understood and on the other hand the resulting measurement variation should be taken into account when using impact values for e.g. funding decisions.
