Is it possible to compare researchers with different scientific interests?

Abstract

The number h of papers with at least h citations has been proposed to evaluate individuals scientific research production. This index is robust in several ways but yet strongly dependent on the research field. We propose a complementary index hI = h2/ N(T)a, with N(T)a being the total number of authors in the considered h papers. A researcher with index hI has hI papers with at least hI citation if he/she had published alone. We have obtained the rank plots of h and hI for four Brazilian scientific communities. In contrast with the h-index, the hI index rank plots collapse into a single curve allowing comparison among different research areas. copyright © 2006 Akadémiai Kiadó, Budapest. All rights reserved.