OpenJournalRank
Compare

What the research says about OJR

OJR (the OpenJournalRank Indicator) is a field-normalized 3-year citation impact computed from OpenAlex, where 1.00 is the world average. It is built from methods that bibliometrics researchers have studied for decades. This article sets what that research says beside each choice we made, notes where the evidence is mixed, and lists what has not been tested yet. It is based on a review of the peer-reviewed literature that we completed in October 2026.

Published Last updated Literature searched through

By OpenJournalRank. OpenJournalRank is run by an independent research group.

How we reviewed. Literature searched through 20 October 2026 in Crossref, OpenAlex, Google Scholar, journal sites and a semantic paper search engine, with AI-assisted searching. Every source was read at least at abstract level, and every DOI, where one exists, was resolved against Crossref or OpenAlex. Numbers are quoted only where we could read them in the source.

Summary

  • The design is established, not new. OJR divides the citations to a journal’s papers by the citations expected for papers of the same subfield, year and document type. That is the classic form of the Leiden journal indicators[1, 2], computed here on open data. It adds safeguards the literature asks for: document-type normalization, no journal self-citations, a top-10% share, bootstrap rank ranges and frozen releases.
  • Expect high agreement with established metrics, but not identity. Published comparisons find strong but imperfect correlations between journal indicators, and weaker ones between field-normalized indicators and the raw Journal Impact Factor[2, 3, 4].
  • Field normalization agrees better with experts. In two national comparisons, SNIP (Source Normalized Impact per Paper) and SJR (SCImago Journal Rank) matched expert journal ratings better than raw impact measures did[5, 6].
  • OpenAlex gives similar ranks but different values. Normalized scores from OpenAlex rank papers and universities much as commercial databases do, but the absolute values differ[7]. OJR ranks should therefore be broadly comparable with other rankings. That evidence comes from paper- and university-level studies; at journal level, our validation study finds a rank correlation of about 0.79 with CWTS SNIP for 2024. OJR values are not interchangeable with Web of Science or Scopus values.
  • OJR has now been compared with SNIP, SJR and expert ratings. When we wrote this review, we found no published study that validates an OpenAlex-based journal indicator against Scopus or Journal Citation Reports metrics. Our validation study has since found rank correlations of 0.79 with CWTS SNIP and 0.78 with SCImago’s SJR for 2024, but both SNIP and SJR agree better than OJR with the Norwegian Register’s expert levels. OJR has not been compared with the Journal Citation Reports metrics.
  • Journal impact says little about a single article. Across 96,031 articles in the UK’s Research Excellence Framework (REF) 2021, journal impact correlated weakly to moderately with the articles’ assessed quality (ρ = 0.11–0.43 by field, where ρ is Spearman’s rank correlation)[8]. OJR describes journals, not papers or people.

How OJR relates to established metrics

OJR compares the citations a journal’s papers received with the citations expected for papers of the same subfield, publication year and document type. A value of 1.00 means the journal’s papers were cited as often as the world average for comparable papers. This “actual over expected” form is the classic Leiden journal indicator, as in the journal-to-field impact score (JFIS)[1] and CPP/FCSm, citations per paper divided by the field’s mean citation score[2].

Two widely used journal metrics share parts of this design. SNIP normalizes for field, uses a 3-year citation window and counts articles, reviews and conference papers, the same document types as OJR[9]. SJR also uses a 3-year window[10]. The Journal Impact Factor (JIF) is not normalized for field. Neither is our Impact factor (2-year), which we show only as context.

How closely should OJR agree with them? Published comparisons show strong but imperfect agreement. Two forms of the same field-normalized journal indicator, a ratio of sums and an average of ratios, correlated at ρ = 0.95 across 8,423 journals[2]. Field-normalized indicators agree less with the raw JIF: SNIP and the JIF correlated at ρ = 0.645 across 9,657 journals[3], and Clarivate’s field-normalized Journal Citation Indicator correlated with the JIF at Pearson r = 0.853[4], a linear correlation rather than a rank correlation. Taking these two figures as rough bounds, we expect OJR to correlate with SNIP at about 0.65–0.85, and at about 0.95 with an average-of-ratios version of itself[3, 11, 4, 2].

Since tested. These were expectations drawn from other indicators. When we wrote this review, we found no published study that had validated an OpenAlex-based journal indicator against SNIP, SJR or the Journal Citation Reports metrics. Our validation study has since compared OJR with SNIP and SJR: ρ is 0.789 with SNIP and 0.782 with SJR for 2024, within the 0.65–0.85 expected for SNIP. OJR has not been compared with the Journal Citation Reports metrics.

Agreement with expert judgement

Citation indicators are often tested against expert ratings of journals. Two large studies favour field normalization:

  • Across about 20,000 journals, SNIP tracked the tiers of the Excellence in Research for Australia (ERA) journal list better than raw impact per paper or the JIF did[5].
  • SNIP and SJR predicted which journals the Norwegian Register places at its top level better than raw impact did[6].

OJR uses the same kind of field normalization. Our validation study compared it with the Norwegian Register: SNIP and SJR separate the register’s top-level journals from the rest better than OJR does, and only part of that gap is explained by meeting abstracts that OpenAlex counts as articles. The ERA 2010 list, where licences allow, would be another benchmark. Like OJR, these expert ratings judge journals, not individual articles.

OpenAlex as the data source

OJR is computed from OpenAlex, an open index of scholarly works released under CC0. Recent studies have compared OpenAlex with Web of Science, Scopus and Dimensions. These are the findings that matter most for OJR.

Ranks agree; values do not

Paper-level field-normalized scores from OpenAlex correlate at ρ = 0.88 with Web of Science and Scopus, and at 0.91 with Dimensions. At university level the correlations are 0.92–0.96, but the concordance of the absolute values with Web of Science is only 0.06[7]. So OJR ranks should be broadly comparable with other rankings, but OJR values should not be read as Web of Science or Scopus values. This evidence comes from papers and universities; journal-level validation is in progress.

Journal coverage

OpenAlex matched 94% of 28,085 active Scopus journals by ISSN, and most of the misses were metadata faults[12, preprint]. OJR ranks only CWTS Core journals for now. CWTS Core is the list of journals behind the Leiden Ranking Open Edition, selected by the Centre for Science and Technology Studies (CWTS) at Leiden University for their international authors and citing base.

Reference coverage

A citation can be counted only if OpenAlex captured the citing paper’s reference list. Per citing paper, OpenAlex records about as many references as Web of Science and Scopus, 34.9 against 33.4, but it matched fewer references in total: 586 million against about 725 million[13]. We publish the share of papers with matched references for every citing year (reference coverage).

Missing references by citing publisher

Reference coverage differs between databases[13]. Our own preliminary checks suggest that in OpenAlex the share of recent papers with no matched references also differs by citing publisher. Those checks were not part of this review, so we do not quote figures here; our reference-gap study reports them, with how they were measured. Normalizing within the same citing year cancels a loss that is the same for every paper, but not one that differs by publisher. Fields where the affected publishers are common, likely including clinical medicine, may be affected most.

Front matter typed as research

Front matter means editorials, letters, book reviews and similar items that are not research papers. One comparison found 303,582 such items in Web of Science typed as articles or reviews in OpenAlex, and checks found Web of Science’s type right 93.5% of the time[14, preprint; 15]. These items add expected citations to the denominator without adding actual citations. See Front matter counted as research under known gaps.

Reviews

Earlier OpenAlex snapshots typed almost no papers as reviews. That has been fixed[16], and review shares now match Web of Science (7.74% against 7.72%)[14, preprint]. Because reviews are cited more, journal pages also show the review share and an articles-only OJR.

Subject classification

OpenAlex is preparing a new topic classifier. Its developers report 81.7% primary-topic accuracy for it, against 24.8% for the current one, and say it would move about 447 million works[17]. These are the vendor’s own figures. The change could reorder subfield baselines and categories between releases.

Non-English and Chinese journals

One study found 14.7% of OpenAlex language tags wrong[18], and another found 37% of Chinese core journals indexed[19, preprint].

A change of data source changes results

When OpenAlex replaced Web of Science or Scopus in an Italian assessment, about 30% of scholars moved by at least one quartile[20]. This is one reason each OJR release is computed from a frozen snapshot and names it.

Each design choice against the evidence

This section judges our own method, so read it with that in mind. We have tried to say plainly where the literature disagrees with a choice or supports it only in part.

In the table, “Supported” means the published evidence backs the choice in general; it does not mean OJR itself has been tested against that evidence. “Mixed” means the literature is divided on the choice, or supports it only in part. Each choice links to where the methodology describes it.

OJR’s design choices, our verdict and the key evidence
ChoiceVerdictKey evidence
3-year citation windowSupportedSJR chose 3 years as the shortest window that covers the citation peak in every Scopus subject area[10], and SNIP uses the same window[9]. Glänzel and Moed call a 3-year window “a good compromise”[21]. Rankings in slow fields such as mathematics and the social sciences are the least stable[22, 23].
Journal self-citations excludedSupported, with caveatsSelf-citation aimed at the impact factor rose after the early 2000s[24], and about 54% of large one-year JIF jumps went with changes in self-citation[25]. Excluding self-citations does not catch citation stacking between journals[26, 27].
Ratio of sums: total actual citations ÷ total expected citationsMixed: contested in the fieldSeveral authors prefer the average of ratios, the mean normalized citation score (MNCS), because it is more consistent[28, 29, 30]; others defend the ratio of sums[31, 32]. Across 8,423 journals the two correlate at ρ = 0.95, but with sharp outliers[2], and they differ more for small units[33]. In favour of our choice, the average of ratios is unstable for the newest papers[2].
Articles, reviews and conference papers, normalized by type, with an articles-only score as wellSupportedSNIP counts exactly these types[9]. Reviews receive 1.34 to 6.74 times the citations of articles, 2.95 times on average[34], and mislabelled reviews are common[35].
Citations from repository-hosted works excludedSupported, with a caveatClarivate also leaves preprint citations out of the JIF[36]. Purchased and anomalous citations have run mostly through some repositories[37]. But citations from preprints do carry real signal[38].
Bootstrap rank ranges, with overlapping ranges read as tiesSupported, with caveatsThe approach matches the Leiden Ranking’s stability intervals[39]. Ranks in the middle of a table are very uncertain[40, 41, 42]. The bootstrap of ranks can be inconsistent near ties[43], and testing whether intervals overlap is conservative, so it declares too many ties[44]. Ranges are computed for the latest year only.
At least 30 papers to be rankedMixed: supported, but at the low endNormalized citation means vary widely up to about 200 papers[45]. At 30 papers, the rank ranges carry most of the work of showing that uncertainty. Separately, we propose flagging journals with fewer than 100 papers as low precision (known gaps).
Quartiles within categoriesSupported, with caveatsIn one study, 10–30% of journals were not meaningfully different from journals across a quartile boundary[46]. The first quartile holds 25–86% of papers, depending on the research area[47], so a first-quartile label covers a very different share of each field’s output.
Top-10% share shown beside the meanSupported, and already shownThe Leiden Ranking made the top-10% share its main indicator after a single highly cited paper lifted one university to 2nd on the mean-based MNCS but 238th on top-10% share[39]. Journal pages, such as CA: A Cancer Journal for Clinicians, show the “Share in the top 10% of their field” and a citation-concentration measure. Neither has an uncertainty interval yet.
Impact factor (2-year) as contextSupported, as context onlyA 5-year window raises the impact factor for about 72% of journals[48], and short windows penalize slow fields[21]. Our Impact factor (2-year) counts journal self-citations, which have been used to raise impact factors[24].

Known gaps and their status

The review points to these gaps. Where work is under way we say so; where it is not, we say what the evidence supports and leave the decision open.

Validation against established metrics

The literature suggests reporting rank correlations overall and by field, agreement with an expert list where licences allow, and a sensitivity check against the average of ratios[3, 11, 4, 2]. Our validation study reports the first two against CWTS SNIP, SCImago’s SJR and the Norwegian Register. It finds broad agreement on ranks, and a gap on the expert levels, where SNIP and SJR both do better than OJR.

Status: Done for SNIP, SJR and the Norwegian Register; the gap on the expert levels is open. Not done for the Journal Citation Reports metrics or the average-of-ratios check.

Front matter counted as research

Editorials, letters and book reviews typed as articles add expected citations to the denominator without adding actual citations[14, preprint]. Counting only papers with at least one matched reference, as the Leiden Ranking does[49], would remove many of them. It might also reduce the effect of reference gaps that differ by publisher; our reference-gap study tests this and finds that, applied to every paper, it also removes most papers of publishers that deposit no references.

Status: Planned. The methodology commits to a fix for document-type problems in the next method version (document types).

Reference gaps that differ by citing publisher

Normalizing within the same citing year does not cancel them, because the loss is not the same for every paper. Our reference-gap study measures them: for most journals the effect is small, and a minority, mostly regional journals cited by publishers that deposit few references, move a lot.

Status: Measured. No correction is applied; the study is repeated with each release.

Precision of small journals and quartile boundaries

Normalized citation means vary widely up to about 200 papers[45], and journals near a quartile boundary often do not differ meaningfully from their neighbours across it[46]. The evidence supports flagging both. For small journals we propose a flag below 100 papers, where scores are least precise. It also supports using pairwise bootstrap probabilities instead of the overlap rule[43, 44], and giving the top-10% share an interval.

Status: Open.

The OpenAlex topic change

When OpenAlex adopts its new classifier[17], subfields and their baselines may shift. The evidence supports reporting how many journals change subfield between releases.

Status: Open.

A policy on citation manipulation

DORA’s recommendation 13 asks metric providers to say what counts as inappropriate manipulation and how they respond[50]. Our responsible-use policy does this: it defines manipulation, including citations arranged between journals, lists the safeguards in place and explains how to report concerns. What is still missing is automatic detection of citation stacking, where journals cite each other unusually heavily[26, 27].

Status: The policy is in place. Automatic detection of citation stacking is open.

What OJR is not

OJR measures the citation impact of a journal’s recent papers, under a published method. A journal’s impact is a weak guide to the quality of any one article in it. In the UK’s REF 2021, the correlation between the quality scores of 96,031 journal articles and the citation impact of the journals they appeared in ranged from ρ = 0.11 to 0.43, depending on the field[8].

In line with DORA[50] and the Leiden Manifesto[51], OJR is not for judging individual articles or the people who wrote them.

By our own reading, the site follows principle 6 of the Leiden Manifesto[51], allowing for differences between fields, and principles 4, 5, 8 and 10 in part. On open data collection and analysis, and letting those evaluated verify them (4 and 5), the input data are open and the method is documented, but others cannot yet rerun the analysis themselves. On false precision (8), ranks come with rank ranges, but not every measure has an interval. On regular review (10), methods are versioned, but there is no fixed review schedule. The site also follows the openness default of the Barcelona Declaration on Open Research Information[52].

Use OJR to compare the citation impact of journals in the rankings. Read ranks with their likely ranges, and quote a value with its release (methodology).

References

Preprints, policy documents and vendor documentation are labelled in this list, and citations of preprints are marked “preprint” in the text.

  1. van Leeuwen TN, Moed HF (2002). Development and application of journal impact measures in the Dutch science system. Scientometrics 53(2):249–266. https://doi.org/10.1023/A:1014808709694
  2. Waltman L, van Eck NJ, van Leeuwen TN, Visser MS, van Raan AFJ (2011). Towards a new crown indicator: an empirical analysis. Scientometrics 87(3):467–481. https://doi.org/10.1007/s11192-011-0354-5
  3. Salvador-Oliván JA, Agustín-Lacruz C (2015). Correlación entre indicadores bibliométricos en revistas de Web of Science y Scopus. Revista General de Información y Documentación 25(2):341–359. https://doi.org/10.5209/rev_RGID.2015.v25.n2.51241
  4. Torres-Salinas D, Valderrama-Baca P, Arroyo-Machado W (2022). Is there a need for a new journal metric? Correlations between JCR Impact Factor metrics and the Journal Citation Indicator—JCI. J Informetrics 16(3):101315. https://doi.org/10.1016/j.joi.2022.101315
  5. Haddawy P, Hassan S-U, Asghar A, Amin S (2016). A comprehensive examination of the relation of three citation-based journal metrics to expert judgment of journal quality. J Informetrics 10(1):162–173. https://doi.org/10.1016/j.joi.2015.12.005
  6. Ahlgren P, Waltman L (2014). The correlation between citation-based and expert-based assessments of publication channels: SNIP and SJR vs. Norwegian quality assessments. J Informetrics 8(4):985–996. https://doi.org/10.1016/j.joi.2014.09.010
  7. Scheidsteger T, Haunschild R, Bornmann L (2025). How similar are field-normalized citation impact scores obtained from OpenAlex and three popular commercial databases? Scientometrics 130:3537–3569. https://doi.org/10.1007/s11192-025-05338-7
  8. Thelwall M, Kousha K, Makita M, Abdoli M, Stuart E, Wilson P, Levitt J (2023). In which fields do higher impact journals publish higher quality articles? Scientometrics 128(7):3915–3933. https://doi.org/10.1007/s11192-023-04735-0
  9. Moed HF (2010). Measuring contextual citation impact of scientific journals. J Informetrics 4(3):265–277. https://doi.org/10.1016/j.joi.2010.01.002
  10. González-Pereira B, Guerrero-Bote VP, de Moya-Anegón F (2010). A new approach to the metric of journals’ scientific prestige: The SJR indicator. J Informetrics 4(3):379–391. https://doi.org/10.1016/j.joi.2010.03.002
  11. Mingers J, Yang L (2017). Evaluating journal quality: A review of journal citation indicators and ranking in business and management. Eur J Oper Res 257(1):323–337. https://doi.org/10.1016/j.ejor.2016.07.058
  12. Alperin JP, Portenoy J, Demes K, Larivière V, Haustein S (2024). An analysis of the suitability of OpenAlex for bibliometric analyses. arXiv. Preprint. https://doi.org/10.48550/arXiv.2404.17663
  13. Culbert JH, Hobert A, Jahn N, Haupka N, Schmidt M, Donner P, Mayr P (2025). Reference coverage analysis of OpenAlex compared to Web of Science and Scopus. Scientometrics 130:2475–2492. https://doi.org/10.1007/s11192-025-05293-3
  14. Mongeon P, Hare M, Riddle P, Wilson S, Krause G, Marjoram R, Toupin R (2025). Investigating document type, language, publication year, and author count discrepancies between OpenAlex and Web of Science. arXiv. Preprint. https://doi.org/10.48550/arXiv.2508.18620
  15. Mongeon P, Hare M, Krause G, Marjoram R, Riddle P, Toupin R, Wilson S (2025). Investigating document type discrepancies between OpenAlex and the Web of Science. Proceedings of the Annual Conference of CAIS. Conference paper. https://doi.org/10.29173/cais1943
  16. Haupka N, Culbert JH, Schniedermann A, Jahn N, Mayr P (2026). Analysis of the publication and document types in OpenAlex, Web of Science, Scopus, PubMed and Semantic Scholar. Quantitative Science Studies 7:179–194. https://doi.org/10.1162/qss.a.406
  17. OurResearch (2026). openalex-walden CHANGELOG.md, “Unreleased” section. GitHub, commit 3a837c0. Vendor documentation. Accessed October 20, 2026. https://github.com/ourresearch/openalex-walden/blob/3a837c0eb020e45ed97ea2a604e568f16413348d/CHANGELOG.md
  18. Céspedes L, Kozlowski D, Pradier C, Holmberg Sainte-Marie M, Shokida NS, Benz P, Poitras C, Ninkov A, Ebrahimy S, Ayeni P, Filali S, Li B, Larivière V (2025). Evaluating the linguistic coverage of OpenAlex. JASIST 76(6):884–895. https://doi.org/10.1002/asi.24979
  19. Zhang L, Cao Z, Liu J, van Eck NJ (2025). Beyond openness: Inclusiveness and usability of Chinese scholarly data in OpenAlex. arXiv. Preprint. https://doi.org/10.48550/arXiv.2512.16339
  20. Abramo G, Cicero T, D’Angelo CA (2026). Data source effects in research performance assessment of individuals and institutions. Scientometrics 131(5):3411–3441. https://doi.org/10.1007/s11192-026-05638-6
  21. Glänzel W, Moed HF (2002). Journal impact measures in bibliometric research. Scientometrics 53(2):171–193. https://doi.org/10.1023/A:1014848323806
  22. Pajić D (2015). On the stability of citation-based journal rankings. J Informetrics 9(4):990–1006. https://doi.org/10.1016/j.joi.2015.08.005
  23. Abramo G, Cicero T, D’Angelo CA (2011). Assessing the varying level of impact measurement accuracy as a function of the citation window length. J Informetrics 5(4):659–667. https://doi.org/10.1016/j.joi.2011.06.004
  24. Chorus C, Waltman L (2016). A large-scale analysis of impact factor biased journal self-citations. PLOS ONE 11(8):e0161021. https://doi.org/10.1371/journal.pone.0161021
  25. Campanario JM (2011). Large increases and decreases in journal impact factors in only one year: The effect of journal self-citations. JASIST 62(2):230–235. https://doi.org/10.1002/asi.21457
  26. Heneberg P (2016). From excessive journal self-cites to citation stacking: Analysis of journal self-citation kinetics in search for journals, which boost their scientometric indicators. PLOS ONE 11(4):e0153730. https://doi.org/10.1371/journal.pone.0153730
  27. Van Noorden R (2013). Brazilian citation scheme outed. Nature 500:510–511. https://doi.org/10.1038/500510a
  28. Waltman L, van Eck NJ, van Leeuwen TN, Visser MS, van Raan AFJ (2011). Towards a new crown indicator: Some theoretical considerations. J Informetrics 5(1):37–47. https://doi.org/10.1016/j.joi.2010.08.001
  29. Lundberg J (2007). Lifting the crown—citation z-score. J Informetrics 1(2):145–154. https://doi.org/10.1016/j.joi.2006.09.007
  30. Opthof T, Leydesdorff L (2010). Caveats for the journal and field normalizations in the CWTS (“Leiden”) evaluations of research performance. J Informetrics 4(3):423–430. https://doi.org/10.1016/j.joi.2010.02.003
  31. Moed HF (2010). CWTS crown indicator measures citation impact of a research group’s publication oeuvre. J Informetrics 4(3):436–438. https://doi.org/10.1016/j.joi.2010.03.009
  32. Vinkler P (2012). The case of scientometricians with the “absolute relative” impact indicator. J Informetrics 6(2):254–264. https://doi.org/10.1016/j.joi.2011.12.004
  33. Larivière V, Gingras Y (2011). Averages of ratios vs. ratios of averages: An empirical analysis of four levels of aggregation. J Informetrics 5(3):392–399. https://doi.org/10.1016/j.joi.2011.02.001
  34. Miranda R, García-Carpintero E (2018). Overcitation and overrepresentation of review papers in the most cited papers. J Informetrics 12(4):1015–1030. https://doi.org/10.1016/j.joi.2018.08.006
  35. Yeung AWK (2019). Comparison between Scopus, Web of Science, PubMed and publishers for mislabelled review papers. Current Science 116(11):1909–1914. https://doi.org/10.18520/cs/v116/i11/1909-1914
  36. Clarivate (2023). Clarivate adds Preprint Citation Index to the Web of Science. clarivate.com. Press release. https://clarivate.com/news/clarivate-adds-preprint-citation-index-to-the-web-of-science/
  37. Ibrahim H, Liu F, Zaki Y, Rahwan T (2025). Citation manipulation through citation mills and pre-print servers. Sci Rep 15:5480. https://doi.org/10.1038/s41598-025-88709-7
  38. Fraser N, Momeni F, Mayr P, Peters I (2020). The relationship between bioRxiv preprints, citations and altmetrics. Quantitative Science Studies 1(2):618–638. https://doi.org/10.1162/qss_a_00043
  39. Waltman L, Calero-Medina C, Kosten J, Noyons ECM, Tijssen RJW, van Eck NJ, van Leeuwen TN, van Raan AFJ, Visser MS, Wouters P (2012). The Leiden Ranking 2011/2012: Data collection, indicators, and interpretation. JASIST 63(12):2419–2432. https://doi.org/10.1002/asi.22708
  40. Greenwood DC (2007). Reliability of journal impact factor rankings. BMC Med Res Methodol 7:48. https://doi.org/10.1186/1471-2288-7-48
  41. Stern DI (2013). Uncertainty measures for economics journal impact factors. J Econ Lit 51(1):173–189. https://doi.org/10.1257/jel.51.1.173
  42. Lyhagen J, Ahlgren P (2020). Uncertainty and the ranking of economics journals. Scientometrics 125:2545–2560. https://doi.org/10.1007/s11192-020-03681-5
  43. Hall P, Miller H (2009). Using the bootstrap to quantify the authority of an empirical ranking. Ann Stat 37(6B):3929–3959. https://doi.org/10.1214/09-AOS699
  44. Schenker N, Gentleman JF (2001). On judging the significance of differences by examining the overlap between confidence intervals. Am Stat 55(3):182–186. https://doi.org/10.1198/000313001317097960
  45. Rogers G, Szomszor M, Adams J (2020). Sample size in bibliometric analysis. Scientometrics 125:777–794. https://doi.org/10.1007/s11192-020-03647-7
  46. Vîiu G-A, Păunescu M (2021). The lack of meaningful boundary differences between journal impact factor quartiles undermines their independent use in research evaluation. Scientometrics 126:1495–1525. https://doi.org/10.1007/s11192-020-03801-1
  47. Miranda R, Garcia-Carpintero E (2019). Comparison of the share of documents and citations from different quartile journals in 25 research areas. Scientometrics 121:479–501. https://doi.org/10.1007/s11192-019-03210-z
  48. Campanario JM (2011). Empirical study of journal impact factors obtained using the classical two-year citation window versus a five-year citation window. Scientometrics 87:189–204. https://doi.org/10.1007/s11192-010-0334-1
  49. CWTS (n.d.). Leiden Ranking Open Edition: Indicators. open.leidenranking.com. Documentation. https://open.leidenranking.com/information/indicators
  50. DORA (2012). San Francisco Declaration on Research Assessment. sfdora.org. Policy document. https://sfdora.org/read/
  51. Hicks D, Wouters P, Waltman L, de Rijcke S, Rafols I (2015). Bibliometrics: The Leiden Manifesto for research metrics. Nature 520:429–431. https://doi.org/10.1038/520429a
  52. Barcelona Declaration on Open Research Information (2024). Barcelona Declaration on Open Research Information. barcelona-declaration.org. Policy document. https://barcelona-declaration.org/

Cite as: OpenJournalRank (2026). What the research says about OJR. Version of October 21, 2026. https://openjournalrank.com/en/research/what-research-says-about-ojr/

To report an error or suggest a correction, email contact@openjournalrank.com.

All research articlesRead the full methodology