How important is choice of the scaling factor in standardizing citations?

Because of the variations in citation behavior across research fields, appropriate standardization must be applied as part of any bibliometric analysis of the productivity of individual scientists and research organizations. Such standardization involves scaling by some factor that characterizes the distribution of the citations of articles from the same year and subject category. In this work we conduct an analysis of the sensitivity of researchers' productivity rankings to the scaling factor chosen to standardize their citations. To do this we first prepare the productivity rankings for all researchers (more than 30,000) operating in the hard sciences in Italy, over the period 2004-2008. We then measure the shifts in rankings caused by adopting scaling factors other than the particular factor that seems more effective for comparing the impact of publications in different fields: the citation average of the distribution of cited-only publications.