Open access (OA) to the research literature has the potential to accelerate recognition and dissemination of research findings, but its actual effects are controversial. This was a longitudinal bibliometric analysis of a cohort of OA and non-OA articles published between June 8, 2004, and December 20, 2004, in the same journal (PNAS: Proceedings of the National Academy of Sciences). Article characteristics were extracted, and citation data were compared between the two groups at three different points in time: at "quasi-baseline" (December 2004, 0-6 mo after publication), in April 2005 (4-10 mo after publication), and in October 2005 (10-16 mo after publication). Potentially confounding variables, including number of authors, authors' lifetime publication count and impact, submission track, country of corresponding author, funding organization, and discipline, were adjusted for in logistic and linear multiple regression models. A total of 1,492 original research articles were analyzed: 212 (14.2% of all articles) were OA articles paid by the author, and 1,280 (85.8%) were non-OA articles. In April 2005 (mean 206 d after publication), 627 (49.0%) of the non-OA articles versus 78 (36.8%) of the OA articles were not cited (relative risk = 1.3 [95% Confidence Interval: 1.1-1.6]; p = 0.001). 6 mo later (mean 288 d after publication), non-OA articles were still more likely to be uncited (non-OA: 172 [13.6%], OA: 11 [5.2%]; relative risk = 2.6 [1.4-4.7]; p < 0.001). The average number of citations of OA articles was higher compared to non-OA articles (April 2005: 1.5 [SD = 2.5] versus 1.2 [SD = 2.0]; Z = 3.123; p = 0.002; October 2005: 6.4 [SD = 10.4] versus 4.5 [SD = 4.9]; Z = 4.058; p < 0.001). In a logistic regression model, controlling for potential confounders, OA articles compared to non-OA articles remained twice as likely to be cited (odds ratio = 2.1 [1.5-2.9]) in the first 4-10 mo after publication (April 2005), with the odds ratio increasing to 2.9 (1.5-5.5) 10-16 mo after publication (October 2005). Articles published as an immediate OA article on the journal site have higher impact than self-archived or otherwise openly accessible OA articles. We found strong evidence that, even in a journal that is widely available in research libraries, OA articles are more immediately recognized and cited by peers than non-OA articles published in the same journal. OA is likely to benefit science by accelerating dissemination and uptake of research findings.
OA articles compared to non-OA articles remained twice as likely to be cited (odds ratio = 2.1 [1.5-2.9]) in the first 4-10 mo after publication
adjusted within-journal association with residual author-selection confounding
Articles published as an immediate OA article on the journal site have higher impact than self-archived or otherwise openly accessible OA articles.
secondary regression favors journal OA, though self-archiving measurement and selection remain imperfect
there seems to be a sustained effect on the absolute number of citations.
the 10–16-month window is too short to establish a sustained long-term effect
OA is likely to benefit science by accelerating dissemination and uptake of research findings.
short-term observational citations cannot identify broader causal scientific benefit
Derived from the full evaluation — not a separate score.
Strengths
Comparing OA and non-OA articles within PNAS holds journal prestige, editorial policy, and broad peer-review context more constant than cross-journal comparisons.
↳ Materials and Methods, Article Cohort
The Introduction identifies self-selection, cross-sectional timing, and journal-level differences as limitations of prior work, then aligns the study design with those gaps.
↳ Introduction, prior-study limitations and final paragraph
The paper discusses contrary findings, skeptical interpretations, preprint timing, institutional repositories, and selective self-archiving rather than presenting the OA advantage as settled.
↳ Introduction, paragraphs 3–8
Limitations
The Conclusions state that OA accelerates scientific advancement and supports APC investment, although the Limitations section concedes that unknown confounders and selection by article importance remain possible.
↳ Discussion, Limitations and Conclusions
The survey of 237 respondents uses self-rated urgency, importance, and quality as a proxy for the principal selection mechanism. It cannot fully establish equivalence between author-selected OA and non-OA articles.
↳ Results, Article and Author Characteristics; Materials and Methods, Article Cohort
The analysis covers one atypical journal and only the first 4–16 months after publication, so transfer to other journals, fields, and long-term citation patterns requires replication.
↳ Discussion, Limitations
The within-journal cohort and adjustment for article and author characteristics make the observed citation association more informative than earlier cross-sectional comparisons. The author survey and self-archiving analysis address selection concerns, but neither fully measures whether authors preferentially purchased OA for work expected to have greater impact. The Abstract and Conclusions then translate the association into claims about accelerated scientific advancement and APC investment that exceed the observational design. Strong contribution and policy relevance therefore coexist with materially constrained craft and contextual positioning.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Sound3.2
Confidence mediumThe within-journal longitudinal comparison meaningfully advances earlier cross-sectional and cross-journal OA studies by holding journal-level conditions constant and separating immediate journal OA from self-archiving. It reduces, but does not resolve, uncertainty about self-selection.
“comparing the bibliometric impact of a cohort of articles from the same journal”
The cohort, repeated citation measurements, covariate adjustment, author survey, and secondary self-archiving analysis form a credible observational strategy. The score is capped because the Conclusions infer accelerated scientific advancement and recommend OA investment despite acknowledged residual selection.
“we were able to statistically control only for known confounders”
The paper follows a coherent progression from the contested literature through crude, adjusted, and secondary analyses. Its final causal and policy language would lead readers beyond the narrower adjusted association established by the results.
“OA is likely to benefit science by accelerating dissemination and uptake of research findings.”
The Introduction engages supportive, contrary, and self-selection-based explanations, and the Limitations section identifies short follow-up, one-journal scope, and residual confounding. The final investment recommendations do not fully preserve those limitations in the stated certainty of the implications.
“There is a possibility of selection bias”
Caveats4 of 4 checks
The reported cohort arithmetic and principal statistical results are internally consistent. A subject-stratified analysis is described but not presented, leaving a minor completeness concern.
The bibliometric component raises no human-subjects issue, but the author survey is reported without an ethics or consent declaration. The supplied article also contains no competing-interest statement despite making policy recommendations about OA.
Flags: 0 declared / 5 total
16 of 16 checkable references verified
20 references in manuscript 4 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 evidence comes from an operating hybrid OA program and measures actual publication choices and subsequent citations. Its observational design makes it informative for policy deliberation but not deployment-ready causal evidence.
“authors could pay US$1,000 if they wanted their article to be immediately OA”
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