Empirical study of the growth dynamics in real career h-index sequences

Based on historical citation data from the ISI Web of Science, this paper introduces a methodology to automatically calculate and classify the real career h-index sequences of scientists. Such a classification is based on the convexity–concavity features of the different temporal segments of h-index sequences. Five categories are identified, namely convex, concave, S-shaped, IS-shaped and linear. As a case study, the h-index sequences of several Nobel Prize winners in Medicine, Chemistry and Economics are investigated. Two proposed factors influencing the growth of the h-index, namely the “freshness” of the h-index core and changes in the rank positions of papers near the h-index point are studied. It is found that the h-index core's “freshness” is particularly relevant to the growth of the h-index. Moreover, although in general more publications lead to an increase of the h-index, the key role is played by those papers near the h-index point.

► We automatically calculate and classify the real career h-index sequences. ► The algorithm is based on the convexity–concavity features of h-index sequences. ► Five categories convex, concave, S-shaped, IS-shaped and linear are identified. ► The h-index core's freshness is particularly relevant to the growth of the h-index. ► Papers near the h-index point play the key role in an increase of the h-index.