Comparative analysis of some individual bibliometric indices when applied to groups of researchers

Abstract

The h-index is a widely used bibliometric indicator for assessing individual scientists or other units of analysis. When evaluating aggregated authors, the h-index may produce rankings that are not consistent with the individual ones. The problem is claimed to affect all h-type indices; while the highly cited publications indicator, which comes from a different class, represents an alternative that is immune to such issue. The main objective of this work is to perform a comparative analysis of some bibliometric indicators originally designed to measure the overall impact of individual scientific production, when applied to the evaluation of groups, to investigate the consistency between the rankings at different levels of aggregation. For that, we use part of a previously reported citation database. The results indicate that, although the consistency at distinct aggregative levels is not formally complied by the h-index and all its variants, it is met with reasonable frequency.