Measuring scientific performance without ties: is scientific leadership the solution?

Abstract

For many decisions in science, evaluators have to select, assess, and rank authors based on their scientific achievements. However, the most diffused scientific performance bibliometric metric &ndash; the <em>h</em> index &ndash; produces many ties, precluding its use to define a full ranking of the authors. In turn, recently, Jorge Hirsch¹ proposes the <em>h_a</em>&nbsp;index (which measures the number of papers of the <em>h</em> core in which the author was the scientific leader) and the associated <em>r_a</em>index(percentage of papers belonging to the <em>h</em> core in which the author was leader) to capture the concept of scientific leadership. We suggest using this last measure to break the ties of the <em>h</em> index. The method is extremely simple and provides a complete solution to this critical problem of the <em>h</em> index. To that end, we develop a two steps procedure, which is able to produce a more granulated ranking of the authors.