nabu.science
MethodologyAboutContact

 

  • Snapshot
  • Key Claims
  • Strengths & Limitations
  • Commentary
  • From the Authors
  • Author Note / Field Feedback
  • Dimension breakdown
  • Impact Profile
Free · About 5 minutes
Report #NBU-02EBF9D9·Framework v5.9·Pipeline v0.4·Evaluated 28 Sept 2026·OECD: 5.8 Other social sciences·Subfield: Statistics, Probability and Uncertainty·Basis: Full text·Type: Cross-sectional study

An empirical analysis of open access citation advantages in library and information science

Sunwoo Lee et al.Sunwoo Lee, Wonsik Shim·2026·Information Research·10.47989/ir31iconf64133 ↗

Primary claim

Non-OA articles received slightly more citations (mean 18.83) than OA articles (mean 17.9), though the difference was minimal.

Supported · the reported means and tiny effect agree, but the aggregate comparison is unadjusted for temporal composition and other confounders
In brief

A useful field-specific map of OA citation patterns whose unadjusted observational design cannot establish that OA choices produce the reported differences.

ReliabilityCaveats4 of 4 checks completeRead the entry ↓
Possible — the claim's weakest requirement is that the effect exists at all. Experimental. Every point below is traceable to a quoted passage, but this forecast has not yet been validated against known replication outcomes.
Abstract

Introduction. This study examines the citation impact of open access (OA) articles in library and information science (LIS). Given limited OA research in LIS compared to other fields, it contributes by examining OA citation advantages over time, journal quartile, OA type, and discipline. Method. Using a dataset of 109,759 LIS journal articles (2001–2024) from WoS, the study classified articles by OA status and type, then applied statistical analyses, including t-tests, ANOVA and time-series analysis. Analysis. Both parametric and nonparametric tests compared citation impacts across OA types, disciplinary scope, publication years, and journal quartiles. Effect sizes were calculated to assess practical relevance. Results. Non-OA articles received slightly more citations (mean 18.83) than OA articles (mean 17.9), though the difference was minimal. Green OA showed the largest citation advantage (28.55 average). OA citation advantages declined from 9.75 citations (2001–2004) to 0.74 (2020–2024). Multi-disciplinary articles fared better than single-discipline articles. Conclusion(s). The study reveals that OA citation advantages in LIS are conditional, varying by OA type, discipline, and period. These findings caution against generalizing the OA citation benefits observed in other fields.

Key Claims

Primary finding · Supported

Non-OA articles received slightly more citations (mean 18.83) than OA articles (mean 17.9), though the difference was minimal.

Possible — the claim's weakest requirement is that the effect exists at all. Experimental. Every point below is traceable to a quoted passage, but this forecast has not yet been validated against known replication outcomes.

the reported means and tiny effect agree, but the aggregate comparison is unadjusted for temporal composition and other confounders

Abstract; Results, Tables 2–3
Quote verified
Secondary · Supported

Green OA showed the largest citation advantage (28.55 average).

OA-type means and pairwise tests agree, but age, self-selection, and OA-designation error remain uncontrolled

Abstract; Results, Tables 4–6
Quote verified
Secondary · Supported

OA citation advantages declined from 9.75 citations (2001–2004) to 0.74 (2020–2024).

period-specific means show the stated decline, although citation exposure is not age-normalised within periods

Abstract; Results, Table 8 and Figure 6
Quote verified
Secondary · Tentative

Interdisciplinary articles benefitted more from OA

unadjusted subgroup differences do not establish an OA benefit or formally test the claimed moderation

Conclusion, p. 1748
Quote verified

Derived from the full evaluation — not a separate score.

Strengths · Limitations

Strengths

#contribution

Large field-specific longitudinal evidence base

The analysis covers 109,759 WoS-indexed LIS articles from 2001–2024 and separates results by OA type, period, disciplinary scope, and journal quartile.

↳ Methods, pp. 1736–1738; Tables 4 and 7–9

#methodological rigour

Effect sizes temper significance claims

Welch tests are paired with nonparametric checks, and Cohen’s d is used to show that several statistically significant contrasts are practically very small.

↳ Methods, p. 1738; Tables 3, 8, and 9

#reporting

Aggregate contradiction addressed openly

The paper explicitly contrasts the overall non-OA mean advantage with the within-period OA advantages and explains the weighting and citation-accrual pattern behind them.

↳ Time-series analysis, pp. 1744–1745; Figure 5

Limitations

#methodological rigour

Central comparisons leave confounding uncontrolled

Separate stratification by period, quartile, and disciplinary scope does not jointly control article age, author and journal characteristics, topic, or self-selection into OA routes.

↳ Methods, pp. 1736–1738; Tables 2 and 4–9

#reporting

Benefit language exceeds observational design

The Conclusion states that interdisciplinary articles “benefitted more from OA,” although Table 7 reports unadjusted subgroup contrasts and no formal interaction or causal identification.

↳ Table 7, p. 1743; Conclusion, p. 1748

#positioning

Consequential limitations remain underdeveloped

The paper does not trace its MIS-category caveat, large Library Journal exclusion, JCR restriction, or OA-classification uncertainty through to the affected conclusions.

↳ Methods, pp. 1736–1737; Discussion and Conclusion, pp. 1747–1748

Does this match your read? Add your assessment ↓

Nabu Commentary

The article makes a genuine descriptive contribution by analysing a large LIS sample across several dimensions that prior field-specific work had not combined. Its statistical reporting is transparent, particularly the use of effect sizes and nonparametric checks alongside mean comparisons. The main constraint is identification: OA groups are compared without joint adjustment for age, journal, author, topic, or self-selection, yet the Discussion and Conclusion sometimes interpret the patterns as benefits or mechanisms. The score also reflects the conflict between Table 8’s significant 2020–2024 result and the Discussion’s statement that statistical validity was ultimately lost.

Evaluate another paper

Free · About 5 minutes · Research preview

Author Note / Field Feedback

Nabu’s assessment, alongside the field’s view.

Are you an author of this paper?

Dimension breakdown

Evidence Score

Out of 5
012345
2.6±0.1

Contribution
3.4

The study adds a large article-level LIS analysis resolved by OA type, disciplinary scope, period, and journal quartile, extending a thin field-specific literature. Its advance remains descriptive and does not explain why the observed differences arise.

Introduction, pp. 1734–1736; Methods, pp. 1736–1738; Tables 4 and 7–9

The study reveals that OA citation advantages in LIS are conditional, varying by OA type, discipline, and period.

Methodological Rigour
2.0

The sample construction, subgroup comparisons, nonparametric checks, and effect sizes are reported transparently. However, the central comparisons do not jointly control article age, journal, author, topic, or OA self-selection, while parts of the interpretation use benefit and mechanism language.

Methods, pp. 1736–1738; Tables 2–9; Discussion and Conclusion, pp. 1747–1748

Interdisciplinary articles benefitted more from OA

Reporting
3.5

The aggregate-versus-period-specific pattern is explicitly identified and explained, and practical effect sizes are distinguished from statistical significance. Reporting is reduced by causal benefit wording and by the statement that significance was ultimately lost despite the significant 2020–2024 result.

Time-series analysis, pp. 1744–1745; Discussion and Conclusion, pp. 1747–1748

ultimately lost statistical validity over time

Positioning
2.6

The paper presents positive, null, and negative findings from the OA citation literature and identifies the most relevant prior LIS study. It does not carry the central confounding limitations or its stated MIS-category caveat through to the practical and interdisciplinary conclusions.

Introduction, pp. 1734–1736; Methods, p. 1736; Discussion and Conclusion, pp. 1747–1748

Any conclusions drawn from this study need to be interpreted accordingly

Lower confidence on Contribution, Positioning — domain match limited.

Reliability

Caveats4 of 4 checks


Internal CoherenceNoted

The tabulated results are broadly internally consistent, but the Discussion misstates the statistical status of the late-period temporal result. This affects interpretation rather than the reported database values themselves.

Research ConductClean

The study uses secondary bibliometric records, declares institutional funding, and makes no data- or code-availability claim requiring verification. No human-subjects ethics declaration or preregistration is required for the reported design.

Flags: 0 declared / 5 total

Reference IntegrityClean

17 of 17 checkable references verified

19 references in manuscript 2 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.

Post-Publication RecordClean

No retraction or expression of concern found in Crossref (includes Retraction Watch).

Last checked: 17 Oct 2026

Sources: Crossref (incl. Retraction Watch) ✓

Impact Profile

Impact areaPolicy and practice · SDG 16
Type of advanceEmpirical
Route to useNamed
AdvanceExtends
CitationsToo early
Policy · patentNot yet traced
ReachedNot reachedNot assessable or unverified
ReadinessValidated
Foundational
Controlled
Piloted
Validated
Implementable
FoundationalImplementable

The primary claim is statistically tested on a large real-world population dataset of LIS articles spanning 2001–2024, which is the population and setting the claim itself concerns.

From the paper
“These results indicate a nuanced approach to OA publications, applicable to both authors and publishers within the LIS discipline.”

The conclusion explicitly names authors and publishers in LIS as the actors who should apply these findings when making OA publication decisions.

Pathway

LIS authors, publishers and institutional policymakers could apply these conditional findings when choosing OA models and setting scholarly communication strategies.

Barrier

Effect sizes are negligible and drawn from a single database in one field, with the authors themselves warning against generalisation, so decision-makers gain little clear guidance.

Evaluated 13 Nov 2026 · GBD 2023 · glm_v1_3_3 · impact_profile_v1_3_3

Report an issue →

AI-generated, human-governed. Something look off? Contact us to request a review.

Keep reading

Related evaluations

Other papers Nabu has evaluated blind, starting with the same field.

  1. Suspected distortion of citations in high-impact cancer journals

    Baptiste Scancar, Jennifer Anne Byrne, David Causeur, et al. · 2026 · Media and communications

    Primary claim

    These findings suggest that paper mills inflate the citation metrics of supported publications and affected journals at scale.

    Tentative
    Evidence Score
    3.5±0.1
    Reliability
    Caveats
  2. The citation trap: Papers published at year-end receive systematically fewer citations

    Chao Ma, Yiwei Li, Feng Guo, et al. · 2019 · Economics

    Primary claim

    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

    Tentative
    Evidence Score
    3.0±0.2
    Reliability
    Caveats
  3. Quantitative Analysis of the Open Access Citation Advantage

    Masashi Kawai, Koichi Ojiro, Makoto Asaoka, et al. · 2022 · Other social sciences

    Primary claim

    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
    Evidence Score
    2.9±0.2
    Reliability
    Caveats

How the rubric works →