Eigenfactor

Eigenfactor
Name	Eigenfactor
Introduced	2007
Developer	University of Washington, Johns Hopkins University
Discipline	Bibliometrics
Frequency	annual
Website	Eigenfactor.org

Eigenfactor

Eigenfactor is a journal-level bibliometric indicator designed to measure the influence of scholarly journals by weighting citations according to the importance of the citing sources. It ranks journals by estimating the share of time that an idealized random reader would spend reading each journal, emphasizing citations from highly cited journals and de-emphasizing citations from lesser-cited journals. The metric complements citation counts and impact measures produced by bodies such as Clarivate, Scopus (Elsevier), and Google Scholar by providing a network-weighted perspective on scholarly influence.

Overview

Eigenfactor quantifies journal influence by modeling citation flows across a network of journals using concepts adapted from network theory and algorithms initially developed for web search ranking. The measure produces two related outputs: the Eigenfactor Score and the Article Influence Score, which together provide journal-level and article-level assessments used by libraries, publishers, and research institutions such as Harvard University, National Institutes of Health, Wellcome Trust, and Max Planck Society. It appears in discussions alongside indicators like the Journal Impact Factor, h-index, and metrics reported by Scimago Journal Rank and Altmetric aggregators.

Methodology

The methodology constructs a directed citation network where nodes represent journals indexed in sources like Web of Science and edges represent aggregated citations between journals during a selected time window. The computation adapts the eigenvector centrality concept from graph theory and the PageRank algorithm introduced by Larry Page and Sergey Brin for Google. A stochastic matrix of citation probabilities is formed and a Markov chain model is iterated until convergence to yield the Eigenfactor Score, while normalization yields the Article Influence Score analogous to the average per-article prestige. Data processing and normalization steps reference databases maintained by organizations such as Institute for Scientific Information and PubMed, and computation often relies on numerical libraries developed in institutions like University of Washington and Johns Hopkins University.

Applications and Use

Libraries and consortia such as Association of Research Libraries and national research funders including National Science Foundation and European Research Council use Eigenfactor-derived indicators to inform subscription decisions, collection development, and program evaluation. Publishers including Elsevier, Springer Nature, and Wiley reference Eigenfactor metrics in journal promotions and portfolio analyses. Academic departments at institutions like Stanford University, University of Cambridge, and University of Oxford may cite Article Influence Scores in tenure reviews and benchmarking exercises, while bibliometricians at Norwegian Centre for Research Data and Leiden University incorporate Eigenfactor outputs into composite indicators and research assessment frameworks.

Limitations and Criticism

Critics from communities including researchers at University of California, Berkeley, University College London, and Brookings Institution have noted sensitivity of Eigenfactor to coverage biases in source databases such as Web of Science and to field-specific citation practices exemplified by journals in Nature Portfolio versus those in specialized outlets. Methodological critiques reference potential distortions introduced by self-citation loops, editorial policies at major publishers like Elsevier or Springer Nature, and the effect of review journals such as Annual Reviews on network centrality. Quantitative analysts from Centre for Science and Technology Studies and civil society groups involved in open science like SPARC have raised concerns about transparency, interpretability for non-specialists, and the risks of using a single metric in high-stakes decisions.

Comparison with Other Metrics

Compared with the Journal Impact Factor produced by Clarivate Analytics, Eigenfactor emphasizes network position rather than per-article citation average; compared with the h-index attributable to Jorge Hirsch, it provides a journal-level prestige measure rather than an author-level productivity metric. Relative to Scimago Journal Rank, which implements a variant of eigenvector centrality on Scopus data from Elsevier, Eigenfactor differs in data sources, normalization choices, and parameterization of the underlying Markov model. Altmetrics providers such as Altmetric.com and Plum Analytics offer complementary attention indicators based on social media and policy mentions, while citation-based indicators from Google Scholar and CrossRef present different coverage and aggregation approaches that affect comparative evaluations.

History and Development

The Eigenfactor project emerged from collaborations among scholars and librarians at institutions including University of Washington and Johns Hopkins University in the early 2000s, drawing on earlier work in network analysis by researchers affiliated with Princeton University and algorithmic innovations by Stanford University computer scientists. Public release and promotion involved partnerships with library consortia and policy stakeholders such as Association of College and Research Libraries and led to adoption in bibliometric studies at bodies like Organisation for Economic Co-operation and Development and UNESCO. Subsequent methodological refinements and software implementations have been discussed at conferences hosted by organizations such as International Society for Scientometrics and Informetrics and in journals published by Elsevier and Springer Nature.

Category:Bibliometrics