Сontemporary Bibliometric Methods for Assessment of Scholarly Publications: Hirsch's Index

Liliya Vulkova, Milena Damianova, Elena Krasteva, Anna Dikovska,
Vladimir Atanasov, and Oleg Yordanov

Abstract

We review the Hirsch's methodology for assessment of scholarly publications. The methodology is centered on an index, known also as h index, which attracted a lot of interest lately. The presentation is intentionally pedestrian making it accessible to readers with basic mathematical and scientific culture only. Using illustrative graphs, we demonstrate that by the virtue of its construction, the h index highly evaluates scientists, who are both productive and conduct high quality research; that is, having high number of publications and enjoying high rate of citations on the majority of their publications. In contrast, the value of h is low for people with a lot of publications, which do not gain much citations and for people whose citations come from few highly cited publications. In this paper, we do not consider in detail several important aspects of the method – among them, the empirically observed linear increase of h with time, the predictive power of h, the scaling necessary if researchers from different disciplines are to be compared, the effect of self-citations on the index, etc. Instead, we direct the readers to suitable references.

In order to assess the availability and applicability of the Hirsch's method to Bulgarian scientific community, we retrieved and analyzed the h index data for all employees of one of local scientific institutions. The institution is selected on the basis of conducting both basic science research and R&D activities. We show how h data can be used to evaluate the promotion policies in the past as well as to suggest measures for improvement. Finally, we study in some detail the subsample of all researchers that undergone a habilitation promotion for the period 2005-2010. For this subsample, we consider other bibliometric indicators: total number of publications, total number of citations and mean number of citations per paper. All these, the h index and the corrected for self-citations h index are calculated at the individual year of habilitation. On this basis we suggest a simple habilitation threshold.