A review of the literature on citation impact indicators

Citation impact indicators nowadays play an important role in research evaluation, and consequently these indicators have received a lot of attention in the bibliometric and scientometric literature. This paper provides an in-depth review of the literature on citation impact indicators. First, an overview is given of the literature on bibliographic databases that can be used to calculate citation impact indicators (Web of Science, Scopus, and Google Scholar). Next, selected topics in the literature on citation impact indicators are reviewed in detail. The first topic is the selection of publications and citations to be included in the calculation of citation impact indicators. The second topic is the normalization of citation impact indicators, in particular normalization for field differences. Counting methods for dealing with co-authored publications are the third topic, and citation impact indicators for journals are the last topic. The paper concludes by offering some recommendations for future research.

💡 Research Summary

The paper provides a comprehensive literature review of citation impact indicators, which have become central to research evaluation, funding decisions, and academic reputation. It begins by outlining the three principal bibliographic databases—Web of Science, Scopus, and Google Scholar—detailing their coverage, metadata quality, and inherent biases. Web of Science offers a curated set of high‑impact journals with precise citation linking but suffers from limited coverage of non‑English and emerging outlets. Scopus expands the journal base and includes broader citation networks, yet its data consistency can vary across fields. Google Scholar aggregates the widest range of scholarly material, including gray literature, but its automatic harvesting leads to duplicate records, non‑scholarly citations, and limited data cleaning, making it less suitable for fine‑grained analyses.

The authors then examine how the selection of publications and citations influences indicator construction. Choices include whether to consider all document types or restrict analysis to original research articles, reviews, or conference papers, and how to define the citation window (e.g., 2‑year, 5‑year, or lifetime). Short windows tend to undervalue recent work, while longer windows may dilute field‑specific dynamics.

Normalization—adjusting for disciplinary citation practices—is identified as the third major theme. Traditional field‑average normalization divides a paper’s citation count by the average citations of its assigned discipline. Recent studies propose more granular approaches: sub‑field normalization, paper‑level reference set normalization, and dynamic clustering that reassigns papers to emergent interdisciplinary groups. These methods aim to mitigate the distortion caused by the growing prevalence of cross‑disciplinary research, where static field classifications become inadequate.

The fourth topic addresses counting methods for co‑authored publications. The “full count” approach assigns a full point to each author, inflating contributions as author numbers rise. Fractional counting allocates 1/n (where n is the number of authors) to each contributor, reducing inflation but ignoring author order and declared contributions. Advanced schemes—weighted fractional counting and order‑based weighting—incorporate information about first‑author, corresponding‑author status, or explicit contribution statements, offering a more equitable representation of individual effort, especially in large collaborations.

Finally, the review surveys journal‑level impact indicators. The Journal Impact Factor (JIF) remains the most widely recognized metric, calculated as the average citations received in a two‑year window, but it is vulnerable to skewed citation distributions, self‑citations, and short‑term bias. Eigenfactor and Article Influence Score extend the analysis by weighting citations according to the prestige of the citing journals, thereby capturing network effects. CiteScore, based on Scopus data, uses a three‑year window and includes all document types, providing a broader perspective. The authors compare these metrics in terms of what they measure (article‑level vs. journal‑level impact), computational complexity, and susceptibility to manipulation.

In the concluding section, the paper outlines several recommendations for future research. First, there is a need for harmonization and integration across bibliographic databases to reduce inconsistencies in citation counts. Second, the development of automated, dynamic normalization techniques that can adapt to evolving interdisciplinary structures is essential. Third, transparent reporting of author contributions should be standardized, enabling more sophisticated counting models. Fourth, journal metrics should move beyond single‑value rankings toward multidimensional dashboards that incorporate openness, societal impact, and citation quality. The authors also suggest leveraging artificial intelligence and real‑time citation monitoring within open‑science infrastructures to produce up‑to‑date, context‑aware impact assessments. Collectively, these directions aim to enhance the reliability, fairness, and interpretability of citation impact indicators in the rapidly changing landscape of scholarly communication.