Does open access mean more citations? A 1,470-journal audit of education research says it depends on the tier

The question: does publishing open access actually get a paper cited more, or is that a comfortable assumption resting on studies that do not stratify by how well-established a journal already was? For over a decade, an 'open-access citation advantage' has been part of the case for OA: make research free to read, and more people can cite it. Camila Lopes Ferreira and colleagues at the Federal University of Technology - Parana (UTFPR), Brazil, tested that claim against a large, single-field dataset rather than a mixed sample spanning disciplines with very different reading and citing habits.

Why it matters: the citation-advantage argument shapes real decisions, from a funder mandating OA publication to an early-career researcher choosing where to submit a paper that needs to be seen. If the advantage is smaller, absent, or reversed in specific fields or specific quality tiers, decisions built on the general claim are built on a claim that does not travel as far as assumed.

What the researchers did. The team analysed 1,470 Education journals indexed in Scopus, examining performance from 2020 to 2023. Journals were split into subscription-based and open-access models, then separately stratified by Scopus's own quartile system (Q1 through Q4) and by membership in the 'top 10%' of journals in the category. Four metrics were compared across these strata: CiteScore, total citations, publication volume, and the percentage of a journal's articles that had been cited at all. The comparisons were tested statistically using t-tests, Mann-Whitney U tests, and chi-square tests, methods suited to comparing two groups (subscription vs. OA) whose underlying distributions cannot be assumed to be normal, which citation counts typically are not.

What they found. Subscription journals dominated the top strata, Q1, Q2, and the top 10%, both in raw quantity (more subscription journals occupied these tiers) and in cumulative citations. Within these top tiers, subscription journals also showed higher CiteScores and a higher percentage of cited articles than open-access journals did. The study's own framing is directional rather than a single effect-size headline: subscription models, in this field and this time window, concentrate more of the citation impact at the top of the distribution than open-access models do.

What this does NOT establish. Bizarus was able to verify this study's abstract, published openly by SAGE alongside the paywalled full text; the full methodology, effect sizes, and discussion of limitations were not accessible to us, and this piece is written from the abstract alone. That is a real constraint, stated plainly rather than papered over: we cannot independently confirm how 'open access' was operationalised (gold, green, diamond, and hybrid OA are different arrangements with different economics, and lumping them together, or not, changes what a finding like this means), nor can we report the actual size of the citation gap the statistical tests detected, only its direction. The finding is also specific to one disciplinary field, Education, over one four-year window; the paper's own reference list includes prior work finding an open-access citation advantage in other contexts (Björk and Solomon's 2012 comparison across fields, for instance), so this result should be read as a field-specific complication of a general claim, not a refutation of it. Most importantly, the study as abstracted describes an association, not a controlled experiment: subscription journals in Education are, on average, older and more institutionally established than the open-access journals that have emerged more recently, particularly since the 2000s. A journal's age and prior prestige plausibly drive both its access model's history and its citation performance, which means the access model itself may be a marker of an older, better-established journal rather than an independent cause of higher citations. The abstract does not describe an approach for isolating the access model's own effect from this confound, and Bizarus did not have access to the full paper to check whether the authors addressed it.

Bizarus interpretation. Read cautiously, this study is a useful corrective to a version of the open-access citation-advantage claim that gets stated more simply than the underlying literature actually supports. Several of the papers this study itself cites, including a 2021 methodological reflection by Basson and colleagues, make a similar point: the citation advantage literature is sensitive to which measures and which subject areas are used, and conclusions drawn from one field or one metric do not automatically generalise. This finding does not mean open access fails to increase readership or citations in general; it means that inside one field, across four specific comparison metrics, established subscription journals currently hold the top citation tiers. Anyone citing 'open access equals more citations' as a settled, field-general fact is overstating what the broader literature, including this new addition to it, actually shows.

What remains unanswered. Whether the same pattern holds outside Education, whether it holds when journal age or founding year is statistically controlled for rather than left as a background factor, and what specific effect sizes the statistical tests in this study actually returned are all questions Bizarus cannot answer from the published abstract alone. A reader who needs the precise numbers, or who needs to check how the authors defined and separated OA sub-types, should read the full paper rather than rely on this summary or the abstract it is based on.

Source record. Ferreira, C. L., Pilatti, L. A., Cantorani, J. R. H., and Picinin, C. T. Citation Metrics of Open Access and Subscription Journals in Education: An Analysis in Scopus's Education Category. Education for Information, Volume 42, Issue 1 (2026). DOI: 10.1177/01678329251364340. Education for Information, Library and Information Sciences category, Q2 (SCImago, 2025 and 2024); the same journal is Q3 in the Education category and Q3 in Information Systems (SCImago, 2025), which is disclosed here because a journal's quartile depends on the category it is being read against, and this piece uses the Library and Information Sciences quartile as the closer match to the paper's own bibliometric method. This piece is based on the article's published abstract; Bizarus did not have access to the paywalled full text and states this as a limitation rather than a formality.