The potential benefit of open access (OA) in relation to citation impact has been discussed in the literature in depth. The methodology used to test the OA citation advantage includes comparing OA vs. non-OA journal impact factors and citations of OA versus non-OA articles published in the same non-OA journals. However, one problem with many studies is that they are small -restricted to a discipline or set of journals-. Moreover, conclusions are not entirely consistent among research areas and 'early view' and 'selection bias' have been suggested as possible explications. In the present paper, an analysis of gold OA from across all areas of research -the 27 subject areas of the Scopus database- is realized. As a novel contribution, this paper takes a journal-level approach to assessing the OA citation advantage, whereas many others take a paper-level approach. Data were obtained from Scimago Lab, sorted using Scopus database, and tagged as OA/non-OA using the DOAJ list. Jointly with the OA citation advantage, the OA prevalence as well as the differences between access types (OA vs. non-OA) in production and referencing are tested. A total of 3,737 OA journals (16.8%) and 18,485 non-OA journals (83.2%) published in 2015 are considered. As the main conclusion, there is no generalizable gold OA citation advantage at journal level.
There is no generalizable gold OA citation advantage at journal level.
broad journal comparisons support the direction, but compositional differences are only partially controlled
In fact, in many areas a gold OA citation disadvantage is observed.
field-level medians repeatedly favor non-OA journals, though the comparisons remain observational
When we consider only the journals of greatest impact (first quartile), the advantage of OA is observed in only one of the 27 areas, whereas the opposite happens in 5 of the 27 areas.
the claim directly reports the six significant Q1 comparisons in Table 6
It seems that the gap between open and non-open access journals decreases over the years when the analysis is controlled by journal age.
small age-stratified groups and an indexing-year proxy limit inference about a narrowing temporal gap
The average impact of a journal is not generally related to the business model used by the publisher.
the broad policy framing exceeds the separately stratified observational comparisons
Derived from the full evaluation — not a separate score.
Strengths
The dataset includes 3,737 OA and 18,485 non-OA journals and reports results across all 27 Scopus subject areas, substantially broadening the scope of earlier discipline-restricted comparisons.
↳ Abstract; Methods, pp. 3–4; Table 1
The paper supplements overall comparisons with SJR-quartile and Scopus-entry-period analyses, showing how prestige and journal age alter the observed OA/non-OA pattern.
↳ Results, pp. 8–9; Tables 6–7
The Scopus coverage limitation is linked to Arts and Humanities, Business, Economics, and Social Sciences and to the possibility that a positive OA effect is not visible in the database.
↳ Conclusions, pp. 11–12
Limitations
Most headline comparisons use separate equality-of-median tests. Quartile and age are examined independently rather than in a joint model that could address several systematic differences between OA and non-OA journals simultaneously.
↳ Methods, pp. 4–5; Tables 3 and 5–7
Gold OA status is assigned through DOAJ membership, with hybrid journals classified as non-OA. The analysis does not quantify misclassification or test whether alternative tagging decisions change the results.
↳ Methods, pp. 3–4
The final subscription, APC-allocation, and publication-strategy recommendations imply operational guidance that was not evaluated through an institutional policy or author-level causal design.
↳ Conclusions, p. 12
The 22,222-journal census and coverage of all 27 Scopus subject areas support a meaningful contribution to the OA citation debate. Median tests are suited to the reported skewed distributions, while the quartile and journal-entry-period analyses help explain some of the overall OA disadvantage. Methodological Rigour is limited because these analyses do not jointly adjust for journal composition, DOAJ-based access classification is not stress-tested, and numerous field-level tests are reported without multiplicity correction. The central journal-level conclusion is more defensible than the final recommendations about publication, subscription, and APC-allocation strategy.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Sound3.5
Confidence mediumThe score reflects a meaningful journal-level extension of the OA citation literature across all 27 Scopus subject areas, supported by a census of 22,222 journals. The advance is primarily broader coverage and stratified description rather than a new identification strategy.
“this paper takes a journal-level approach to assessing the OA citation advantage”
Median-based comparisons suit the reported skewed distributions, and quartile and Scopus-entry-period analyses probe prestige and age. The comparisons remain largely unadjusted, the DOAJ classification is not validated through sensitivity analysis, and multiple field-level tests are not corrected.
“the characteristics of each of these sources might influence the results”
The paper follows its three research questions through dedicated results and retains the journal-level qualifier in its main conclusion. Explanatory mechanisms and recommendations in the closing discussion are stated more confidently than the descriptive comparisons warrant.
“there is no generalizable gold OA citation advantage at journal level”
The discussion engages findings both supporting and questioning an OA citation advantage and distinguishes journal-level gold OA from article-level and green OA evidence. The Scopus coverage limitation is traced to specific fields and to the possibility that a positive effect may be hidden.
“there may be a positive effect of OA upon citation impact, but it is not visible in the database used”
No concerns4 of 4 checks
Sample totals and central findings are internally consistent with the reported tables. No material statistical impossibility or methods–results contradiction was established.
The study uses aggregate bibliometric metadata and does not require human-subject ethics review. No unsupported data, code, ethics, or conflict-of-interest availability claim is made.
Flags: 0 declared / 5 total
26 of 26 checkable references verified
27 references in manuscript 1 have no canonical index record — counted, but not index-checkable 1 reference confirmed by manual review
References not verified automatically are resolved manually within 24 hours.
No retraction or expression of concern found in Crossref (includes Retraction Watch).
Sources: Crossref (incl. Retraction Watch) ✓
Where this paper’s evidence sits on the path from initial observation to real-world use.
The study uses large-scale real-world bibliometric data, but it evaluates no implemented policy, institutional decision rule, or intervention. Readiness therefore remains at the descriptive evidence stage.
“The DOAJ list was used as the source for OA tagging.”
AI-generated, human-governed. Something look off? Contact us to request a review.
Keep reading
Other papers Nabu has evaluated blind, starting with the same field.
Baptiste Scancar, Jennifer Anne Byrne, David Causeur, et al. · 2026 · Media and communications
These findings suggest that paper mills inflate the citation metrics of supported publications and affected journals at scale.
TentativeChao Ma, Yiwei Li, Feng Guo, et al. · 2019 · Economics
papers published between October and December on average get as much as 18.5% fewer citations than those published in the other months in the year
TentativeMasashi Kawai, Koichi Ojiro, Makoto Asaoka, et al. · 2022 · Other social sciences
日本の機関リポジトリによるオープンアクセス論文の数は限られるものの，その被引用数と様々なタイプのオープンアクセス論文の被引用数の比較からは，日本の機関リポジトリによるオープンアクセスのメリットは確認できなかった．
Tentative