SCImago Journal Rank

• 1. Introduction

  ›

  SCImago Journal Rank asks a deceptively simple question: when a scientist cites a journal, does every citation count the same? The answer, according to the researchers behind SJR, is no. A citation from Nature carries a different weight than a citation from an obscure proceedings volume, and ignoring that difference distorts our picture of scientific prestige entirely.

  The metric was developed by the Scimago Lab, which originated from a research group at the University of Granada. The name itself is a small curiosity. SCImago is not an acronym, though the first three capitalized letters do stand for SCientific Influence. The suffix "mago" appears to have been chosen at random.

  What unfolds from that origin is a ranking system built on network logic, borrowed from the same mathematical world that powers web search. How the algorithm works, what it reveals about scientific publishing, and where it contradicts our intuitions about journal quality are the questions this documentary will follow.

• 2. Citations as Endorsements

  ›

  Prestige, in the SJR framework, is understood as two things working together: the number of endorsements a journal receives and the prestige of whatever is doing the endorsing. A citation is treated as an endorsement, and not all endorsements carry equal weight.

  This view shapes how the SJR indicator assigns its values. Citations from highly-ranked journals pass along more prestige than citations from lesser-known ones. The logic is straightforward: if respected researchers in respected journals keep pointing back to a particular publication, that pattern says something meaningful.

  Yet the source of this metric also notes a complicating finding. Studies of methodological quality and reliability have found that "reliability of published research works in several fields may be decreasing with increasing journal rank." That runs against widespread expectations. The assumption that a higher-ranked journal reliably publishes more trustworthy work turns out to be contested by the evidence.

• 3. The Algorithm Behind the Score

  ›

  SJR is a variant of the eigenvector centrality measure used in network theory. That class of measures establishes the importance of a node in a network based on the principle that connections to high-scoring nodes contribute more to the node's own score. The SJR indicator was designed specifically for extremely large and heterogeneous journal citation networks.

  The computation begins by assigning an identical amount of prestige to every journal. Then an iterative procedure redistributes that prestige, with journals transferring their accumulated standing to each other through citations. The process continues until the difference between prestige values in consecutive iterations no longer reaches a minimum threshold.

  The calculation runs in two phases. First, the algorithm computes a size-dependent measure called the Prestige SJR, which reflects each journal's total prestige. Second, that figure is normalized into a size-independent measure, the SJR indicator itself. The end result represents the average number of weighted citations received during a selected year per document published in that journal during the previous three years, as indexed by Scopus. The SJR indicator is free to access, and its algorithm is inspired by PageRank.

• 4. SJR Versus Its Rivals

  ›

  The SJR indicator is closely related to the Eigenfactor score, and the two share a similar calculation approach. The difference lies in their data sources: SJR draws on the Scopus database, while Eigenfactor relies on the Web of Science database. Other differences exist between them as well.

  Beyond Eigenfactor, SJR also sits alongside the traditional impact factor, which measures average citations per document. The SJR platform provides its own direct alternative to the impact factor: average citations per document over a two-year period, abbreviated as Cites per Doc. (2y).

  The SCImago website publishes both the SJR and the H index, and the results between those two metrics diverge sharply. In the SJR ranking, Nature sits at number one, Science at number two, and the New England Journal of Medicine at number three. The H index produces an entirely different ordering: Ca-A Cancer Journal for Clinicians ranks first, Foundations and Trends in Machine Learning ranks second, and Nature Reviews Molecular Cell Biology ranks third. The gap between rank 1 and rank 621 for the same journal, Nature, across the two metrics, shows how differently each system defines influence.

• 5. What Prestige Per Article Means

  ›

  SJR is described as a size-independent indicator, which means a journal's score is not inflated simply by publishing a high volume of articles. Its values order journals by their "average prestige per article," making it applicable for journal comparisons in science evaluation processes.

  That design choice has practical consequences for how institutions and funding bodies assess journals. A large journal that publishes thousands of articles but receives only modest citations per piece will not outrank a smaller journal whose fewer articles attract citations from highly prestigious sources.

  The iterative steady-state solution the algorithm reaches is what makes the comparison coherent across fields. Because the process is developed to handle heterogeneous citation networks, journals from very different disciplines can be placed on a common scale, even though citation practices vary widely across the sciences.