Abstract:
In this paper, we look at the issue of the high-end of research performance
which is captured in the tail of a citation distribution. As the mean is insufficient to capture
the skewness of such distributions, a consistency or concentration measure is the additional
parameter needed. We show that the h-index is only approximately a heuristic mock of a
composite indicator built from three primary indicators which are the number, mean and
consistency term. The z-index is able to sense the change in consistency in the distribution
due to the outliers in the tail of the distribution.
