Exploring the limitations of the h-index and h-type indexes in measuring the research performance of authors

With the introduction of an increasing number of evaluation indexes, researchers have begun to pay attention to the limitations of such indexes in research evaluation, understanding which to avoid misusing and making evaluation more scientific and reasonable. Analysing the principles of the h-index, g-index, AR-index, p-index, integrated impact indicator (I3), and academic trace, this paper explores their limitations in measuring the research performance of authors from the perspectives of consistency, the degree of discrimination, and the statistical relationship between the values of indicators and the number of publications and citations. There are some interesting findings. These six indicators are highly consistent, and they are all more susceptible to the number of publications than to the frequency of citations. Among them, the h-index has the lowest degree of discrimination, followed by the g-index, I3, AR-index, p-index, and academic trace. The g-index ignores papers and citations other than the g-core. Moreover, compared to the h-index, the accumulation of citations makes it easier for the g-index to be equal to the number of papers published by an author, and once its value equals the number of papers, subsequent citations received by these papers will no longer contribute to the growth of the g-index unless the author publishes a new paper. Additionally, the AR-index ignores the h-tail papers and citations, which underestimates the impact of many researchers. Moreover, the p-index is insensitive to highly cited papers. Furthermore, the I3 is very vulnerable to the influence of the extremums in a data set. Finally, we propose considerations and suggestions for the research performance evaluation of authors.