The global disruption index (GDI): an incorporation of citation cascades in the disruptive index

The disruption index (DI) proposed by Funk and Owen Smith (2017), and improved by Wu et al. (2019) has been widely studied and applied in bibliometrics recently. This paper examines the relationship between disruptiveness and influence and finds that highly influential publications tend to be more disruptive. However, the original formula of the disruption index (DI) shows inaccurate results in our sample data. We identify three limitations of DI and propose the global disruption index (GDI), which considers multiple generations of cited or citing documents in citation cascades, to address these limitations. Our empirical study, based on a dataset of 134,003 papers published in the field of Information Science Library Science from 1900 to 2021, and all articles published in Scientometrics, shows that GDI has a significant correlation with the citation count and the number of disruptive citations. We argue that GDI is a more appropriate metric for measuring the disruptiveness of scientific publications and has great significance for academic evaluation and scientific development in the future. Furthermore, our study suggests that the number of consolidated citations also facilitates publications’ disruptiveness. These findings have important implications for understanding the nature of scientific impact and for developing more accurate and comprehensive bibliometric measures.