Analysis of journal rankings confirms that more cited articles contain more references

Scientific articles are used to be cited according to various factors, such as their quality, novelty, methodology, etc. There are also other factors which were identified to influence the citation impact of articles, such as the length of the articles, the number of authors, and the number of references etc. The influence of the number of references was analysed in this study. Journals from various research fields, such as Chemical Engineering, Mathematics, Computer Science, Neuroscience, and Psychology, were ranked according to their SCImago Journal Rank (SJR) values into deciles for the years 2000, 2005, 2010, 2015, 2020, and 2023. The number of references related to one article was calculated for each decile and then correlated against the decile ranking (from 1 to 10). Significant correlation coefficients ranging from -0.880 to -0.982 were obtained. This approach is different from the common approach to analyse the citations and references of the individual articles. The negative slopes of such linear plots demonstrate that the numbers of references are related to the citation impact of journals and, in turn, articles published by them. This is in line with the journal-article-level Matthew effect so that the authors prefer to read the higher-ranked journals and then cite their articles.