Normalization of h-indexes
Reports on a project by an Indiana group to correct h-index rankings so as to be able to compare people across disciplines, by the obvious method of normalization.
However, it fails to address the well-known fault of the h-index: it gives lower rankings to those who publish few, but very important, papers, than it gives to those who publish many, though far less significant papers. Thus for example, Feynman doesn't make it into their top 100 physicists, or mathematical physicists, or theoretical physicists. Nor does Woodward make it into the top 100 organic chemists. (Though it is still in beta, which may distort the results).