The OA citation advantage is confirmed
unadjusted self-selection prevents the observational comparison from confirming an OA effect
OA papers not only have the great advantage of total downloads, but also have the feature of keeping sustained and steady downloads for a long time
Figure 6 supports the pattern descriptively, but unequal follow-up and uncontrolled article differences limit inference
OA articles attract a litter more social media attention than non-OA articles
group means consistently support the modest wording, though the comparison remains unadjusted
Derived from the full evaluation — not a separate score.
Strengths
Figure 6 separates accumulated page views into early, middle, and late periods and reports markedly different middle-period slopes for OA and non-OA articles. This adds a temporal perspective beyond static citation counts.
↳ Results, Dynamic comparison; Figure 6
OA and non-OA articles are drawn from Nature Communications during the same publication window, reducing differences attributable to journal brand and platform.
↳ Data and Method
The Introduction cites both studies supporting an OA advantage and skeptical or qualified findings, including randomized evidence. The paper does not portray the underlying literature as uniformly settled.
↳ Introduction
Limitations
The paper asks whether authors selected their best articles for OA but performs no matching, regression adjustment, or sensitivity analysis. The observed group differences therefore cannot identify an effect of OA status.
↳ Data and Method; Limitation
The Abstract and Conclusion say the OA advantage is “confirmed,” although the analysis consists of uncontrolled descriptive comparisons. This wording invites a stronger causal interpretation than the evidence supports.
↳ Abstract; Conclusion and discussion
The sample covers one journal and nature.com page views, excluding usage through personal sites, repositories, and other publisher systems. The paper does not test relevant boundary conditions across settings.
↳ Data and Method; Limitation
The paper's strongest addition is the temporal analysis in Figure 6, which provides a descriptive view of how platform attention accumulates after publication. The same-journal sample improves comparability, but the Data and Method section reports no analytical control for the selection and author-level factors later identified in the Limitation section. Because the Abstract and Conclusion call the OA advantage “confirmed,” the inferential language exceeds what those unadjusted comparisons can establish. Balanced engagement with prior literature and a specific limitations discussion support the Positioning and Reporting scores, but neither offsets the central methodological constraint.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Limited2.6
Confidence highThe study adds a same-journal comparison across citations, platform views, and social-media attention, with the accumulated-view analysis providing its clearest incremental contribution. The central OA-advantage finding otherwise largely confirms an established literature.
The slope of the trend line of the OA curve is 7.87, when the slope of the non-OA curve is only 1.74.
The sample and descriptive comparisons are clearly defined, but OA status is self-selected and no matching, covariate adjustment, uncertainty analysis, or sensitivity analysis addresses that confounding. Unequal follow-up and changing sample composition also constrain the dynamic comparison.
did the authors of the OA articles intentionally select their best articles for free access?
The progression from research questions to static and temporal comparisons is easy to follow, and the figures map directly to the reported outcomes. The Abstract and Conclusion nevertheless describe the OA advantage as confirmed without retaining the paper's own confounding caveats.
The OA citation advantage is confirmed
The Introduction presents supportive and skeptical OA literature, including randomized evidence, and the Limitation section identifies self-selection and platform-bounded measurement. Those limitations are not traced into the certainty of the concluding confirmation claim.
Many variables need to be considered.
Lower confidence on Positioning — domain match limited.
Caveats4 of 4 checks
A denominator typo affects the reported non-OA sample size in the citedness paragraph, although the surrounding sample totals and calculation make the intended value unambiguous.
The analysis uses publicly displayed article-level metrics, and the supplied text presents no design-triggered ethics issue or contradictory conduct declaration. Absence of preregistration, formal data sharing, and code sharing is not treated as a conduct concern for this descriptive bibliometric study.
Flags: 0 declared / 5 total
22 of 22 checkable references verified
25 references in manuscript 3 have no canonical index record — counted, but not index-checkable 2 references 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).
Correction published 04 Apr 2015 · listed in Crossref
Sources: Crossref (incl. Retraction Watch) ✓
Where this paper’s evidence sits on the path from initial observation to real-world use.
The work is an initial observational comparison rather than validated policy guidance or an implementation study. Unaddressed selection into OA prevents direct translation of the reported differences into an intervention claim.
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