Distinct clusters of CiteScore and percentiles in top 1000 journals in Scopus

Abstract

CiteScore, Scopus/Elsevier’s open journal metric, is an attractive alternativeto Clarivate Analytics’ impactfactor. Inmid-2020, theequation used to calculate the CiteScore changed, reflecting a four-year window of data versus a previous three-year data set. Extrapolating CiteScore data from Scopus for the top 1000 ranked journals, we wanted to appreciate how CiteScore trended over time. We found that, on average, CiteScore increased consistently each year between 2015 and 2019, from 13.877 to 16.536. Broadly, this reflects a greater number of citations per publication over time, so a constant rise in citation rate. Academics should not erroneously mistake this rise as a higher level of quality. In addition, k-mean clustering of the percentile and CiteScore showed the existence of three distinct clusters for the top 1000 ranked journals, which aggregated together due to their distinct similarities (similar mean). This pattern may assist researchers to study how the pattern of the distribution of CiteScore and percentile changes over time, and monitor how the CiteScore methodology has evolved over the years.