Pith. sign in

REVIEW 4 major objections 5 minor 7 references

International collaboration of Ukrainian scholars: Effects of Russia's full-scale invasion of Ukraine

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read After Russia's full-scale invasion, Ukraine's universities—not its national academy—gained the most from international collaboration.

desk verdict First solid descriptive account of how the full-scale invasion reshaped Ukrainian collaboration; the sector-gains claim needs a direct interaction test and robustness to compositional change. read the letter →

arxiv 2505.20944 v2 pith:W5RAZIB3 submitted 2025-05-27 cs.DL

Myroslava Hladchenko This is my paper
classification cs.DL
keywords internationalcollaborationUkrainianscholarscitationimpactRussiainvasionofUkraineNationalAcademySciencesuniversitiesco-authorshipscientometrics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether Russia's full-scale invasion of Ukraine changed how Ukrainian scholars collaborate internationally, comparing the two years before (2020–2021) with the two years after (2022–2023). Using publication records of articles with at least one Ukrainian-affiliated author, it finds that universities increased their share of internationally co-authored articles from 26.2% to 32.3%, while the share for the National Academy of Sciences of Ukraine stayed near 40%. The citation impact of internationally co-authored articles rose statistically significantly for universities but did not change for the academy. The paper reads this as evidence that universities gained more from international collaboration during the war, even though both sectors remain far below the 60–70% international-collaboration share typical of developed countries. The finding matters because it suggests where support for Ukrainian research during and after the war could do the most good.

What carries the argument

The analysis runs on a dataset of 61,424 articles with at least one Ukrainian affiliation published 2020–2023, split by authorship pattern into internationally co-authored, domestically co-authored, and single-authored articles; articles by scholars with foreign co-affiliations were excluded. The central measurement is the field-normalised citation impact (FNCI), which divides each article's citation count by the expected count for its year, discipline, and publication type, alongside the percentage of articles in the global top 10% of citations. The statistical workhorse is negative binomial regression, used because citation impact data are overdispersed; separate models test whether the 2022–2023 period changed citation impact, whether universities differ from the NASU, and how disciplines differ. The concept of hyperauthorship—very long author lists typical of large physics and medicine collaborations—frames the auxiliary finding that adding authors or foreign-affiliation countries raises citation impact in some disciplines and for universities more than for the NASU.

What would settle it

Re-run the analysis on the fixed set of scholars who published in both 2020–2021 and 2022–2023 and recompute the university share of internationally co-authored articles and their citation impact. If the share no longer rises from 26.2% to 32.3% or the citation impact increase disappears, the paper's central claim—that universities gained more from international collaboration because of the war—would collapse.

Watch

Extended reading notes

Core claim

The paper's central discovery is a divergence between the two main Ukrainian research sectors during the war. In 2020–2021, the NASU produced about 40% internationally co-authored articles; in 2022–2023 this share stayed roughly unchanged. Universities, starting lower, raised their share from 26.2% to 32.3%, with a corresponding shift away from single-authored and domestic-only articles. The statistically significant increase in field-normalised citation impact occurs only for universities, and the university citation advantage over the NASU grows in 2022–2023. Russia dropped from first to ninth among collaboration partners for the NASU and from first to fourth for universities, with Poland, Germany, the USA, China, and the Czech Republic absorbing much of the shift. Disciplinarity matters: most NASU international articles are in physical sciences and engineering, but their citation advantage over universities in that field nearly vanished in 2022–2023, while universities consistently lead in biomedical and health sciences. The paper concludes that international collaboration is a viable lifeline for Ukrainian scholars, and that universities in particular converted collaboration into citation impact.

Load-bearing premise

The load-bearing premise is that comparing 2020–2021 with 2022–2023 isolates the effect of the invasion, even though over the same period the population of Ukrainian scholars fell by 22%, with 6.3% emigrating, 5.5% internally displaced, and 1,518 joining the army; if the scholars who left were less collaborative, the university gains could be an artifact of who remained.

Editorial extensions

If this is right

  • If universities genuinely convert international collaboration into citation impact better than the academy does, war-era funding for international co-authorship should be aimed at university researchers to maximize research-quality returns.
  • The academy's flat share and flat citation impact imply that the same policy instruments will not lift both sectors; NASU institutes may need discipline-targeted collaboration support, especially in physical sciences and engineering.
  • Russia's drop from first to ninth (NASU) and from first to fourth (universities) among collaboration partners is a structural reorientation of Ukrainian publication networks toward the EU that is likely to persist after the war.
  • With both sectors at 32–41% internationally co-authored share versus 60–70% in developed countries, increasing international collaboration remains a concrete, measurable lever for raising Ukraine's research visibility.
  • The growing citation advantage of university international articles means domestic-only and single-author output is falling further behind, widening the gap in top-10% most-cited shares.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The 2020–2021 versus 2022–2023 comparison does not adjust for the 22% drop in scholar numbers, emigration, displacement, or military service; if less-collaborative scholars left the population, part of the university gains could reflect who remained rather than changed behavior. This is my inference, not a claim the paper makes.
  • A balanced-panel test—restricting to scholars who published in both periods—would separate composition effects from behavior effects; if the university citation gain vanishes, the central conclusion would need revision.
  • The paper reports that high mean citation impact in biomedical and health sciences and social sciences is driven by outliers, with medians below 0.5; robustness checks on medians or trimmed means could reveal whether the university advantage is broad-based or concentrated in a few highly cited papers.
  • The stronger author-count/citation relationship for universities suggests that policies encouraging large multinational teams would mainly help universities; for the academy, discipline-specific funding for physical sciences may matter more than generic collaboration promotion.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper examines the effects of Russia's full-scale invasion of Ukraine on international collaboration of Ukrainian scholars affiliated with the NASU and universities. Using Scopus data from 2020 to 2023, it compares the share of internationally co-authored articles, the top collaborating countries, disciplinary patterns, and citation impact (FNCI) before (2020-2021) and after (2022-2023) the invasion. The central claim is that universities gained more from international collaboration than the NASU: the international-collaboration share increased from 26.2% to 32.3% for universities while remaining around 40% for the NASU, and the citation impact of internationally co-authored articles increased significantly for universities (Table 1, Model 2) but not for the NASU (Model 1). The paper also documents a shift away from Russian co-authorship and reports discipline-specific regression analyses. The study is descriptive and uses negative binomial regressions appropriate for overdispersed citation counts, but several inferential gaps affect the headline comparison.

Significance. This is a timely, policy-relevant bibliometric study of a scientific community under war conditions, using a large CWTS in-house Scopus dataset and field-normalized citation indicators. The paper's strengths include the comprehensive coverage of all Ukrainian-affiliated articles, the attention to hyperauthorship, the breakdown by discipline, and the explicit recognition of the severe contraction of the Ukrainian scholarly population. If the main findings survive additional robustness checks, the paper provides valuable evidence on how international collaboration can support a research system under extreme stress. However, the central claim that universities gained more than the NASU rests on a pre/post comparison without adjustment for compositional changes in the publishing population and without a formal sector-by-period interaction test. These are correctable issues rather than fundamental flaws, so the paper is potentially suitable for publication after major revision.

major comments (4)
  1. [Introduction; Materials and methods] The central comparison of 2020–2021 with 2022–2023 is estimated on the observed set of published articles, but the publishing population changed dramatically: as reported in the Introduction, the number of scholars fell by 22%, 6.3% emigrated, 5.5% were internally displaced, and 1,518 joined the army. If scholars with existing international ties were differentially retained or able to publish, the increase in universities' international-collaboration share (26.2% to 32.3%) and the positive 2022–2023 coefficient in Table 1, Model 2 could be driven by compositional selection rather than by a change in collaboration behavior. Please add author-level fixed effects, a balanced-author restriction, or pre-war publication-propensity weighting, and discuss how selection affects the interpretation.
  2. [Statistical tests, Table 1] The claim that universities gained more from international collaboration than the NASU is not directly tested: Models 1 and 2 estimate sector-specific period effects, while Models 3 and 4 estimate the university advantage within each period. No test of a sector×period interaction is reported, so the null hypothesis of parallel period shifts between sectors is never formally evaluated. Report a pooled model with an interaction term (and its confidence interval) or an equivalent test.
  3. [Statistical tests, Tables 1–3] No confidence intervals or effect-size measures are reported for any regression coefficient; the deviance explained values are extremely small (e.g., 5.28e-05 in Model 1 and 0.0017 in Model 2 of Table 1), and the single McFadden pseudo-R² entry (0.0773) in Table 1 appears misaligned with the models. With over 7,000–22,000 observations, statistical significance at the 0.001 level may reflect sample size rather than a substantive citation impact. Please report exponentiated coefficients with confidence intervals and interpret the magnitude.
  4. [Materials and methods] Articles co-authored by both NASU and universities are counted once in each category, making the two sector samples non-independent and the sector percentages computed over overlapping numerators. Because the overlap between the two sectors may have changed between 2020–2021 and 2022–2023, this double counting can bias the comparison of sector-specific trends. Please assess the sensitivity of the main results to assigning such articles to a separate joint category or to both with appropriate clustered standard errors.
minor comments (5)
  1. [Results, Figure 1 and elsewhere] The decimal comma in '32,2%' is inconsistent with the decimal points used elsewhere in the manuscript; please standardize the notation.
  2. [Results, Figure 9 text] There is a typo in 'univerisities' in the paragraph referring to Figure 9; please correct it.
  3. [Results, Figures 5–6] The descriptions of trend lines in Figures 5 and 6 report only qualitative impressions ('positive slope', 'weaker correlations') without reporting the underlying correlation coefficients or slopes; please add these values.
  4. [Statistical tests, after Table 3] The sentence 'Model 3 specifically examines the relationship between authorship patterns and citation impact' should explicitly say 'Table 3, Model 3' to avoid confusion with Table 1.
  5. [Materials and methods] The negative binomial regressions are described only briefly; please state explicitly which covariates are included in each model and which software was used for estimation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the study's estimates are empirical findings from bibliometric data, not derivations that reduce to their own inputs.

full rationale

This paper is an empirical bibliometric analysis. The central quantities—shares of internationally co-authored articles, top collaborating countries, FNCI values, and negative binomial regression coefficients—are all estimated directly from Scopus data, and the paper does not claim to derive them from first principles or from a fitted parameter that is then renamed as a prediction. The author's earlier works (Hladchenko et al., 2020; Hladchenko & Westerheijden, 2018) are cited only as background on Ukrainian higher education policy and institutional history; they do not supply the load-bearing premise for any statistical claim. The headline inference that universities gained more from international collaboration than the NASU is a comparison of estimated period and sector coefficients, which is testable against the data rather than being true by construction. The concern raised by a skeptical reader—that compositional shifts in the publishing population (22% decline in scholars, emigration, displacement, army service) could confound the before/after comparison—is a substantive threat to causal validity, but it is not circularity: the estimates remain data-derived and could in principle go in either direction. No self-definitional step, fitted-input-renamed-as-prediction, imported uniqueness theorem, or ansatz-smuggled-via-citation appears in the manuscript. The analysis is therefore not circular; any weakness lies in inference and external validity, not in the derivation chain.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No hand-fitted constants are introduced; regression coefficients and dispersion parameters are estimated from the data and are the object of study, not inputs to a derivation. The main assumptions are data reliability, FNCI validity, and the comparability of the two periods despite changing researcher composition.

assumptions (5)
  • domain assumption The CWTS in-house Scopus database provides complete and accurate coverage of articles with Ukrainian affiliations.
    All analyses rely on this dataset (Materials and methods, first paragraph).
  • domain assumption Field-normalised citation impact (FNCI) computed by CWTS is a valid measure of citation impact across disciplines.
    Used throughout to compare citation impact; FNCI normalization is standard in bibliometrics.
  • domain assumption Authorship and affiliation data accurately reflect collaboration.
    The study infers collaboration from co-authorship and counts articles with foreign affiliations as internationally co-authored.
  • ad hoc to paper Articles co-authored by both NASU and universities can be counted once in each category without biasing comparisons.
    The authors state 'Articles co-authored by scholars affiliated with both the NASU and universities were counted twice' (Materials and methods), but no sensitivity analysis is provided.
  • ad hoc to paper The pre/post comparison isolates the effect of the invasion.
    The paper notes a 22% drop in scholar numbers but does not control for researcher emigration, military service, or database coverage changes (Introduction, State Statistics Service paragraph).

how reviews work

0 comments
Cite this review

Pith. "Pith review of International collaboration of Ukrainian scholars: Effects of Russia's full-scale invasion of Ukraine." pith.science (2026). https://pith.science/paper/W5RAZIB3

@misc{pith2026250520944,
  author       = {Pith},
  title        = {Pith review of: International collaboration of Ukrainian scholars: Effects of Russia's full-scale invasion of Ukraine},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W5RAZIB3}},
  note         = {Machine review of arXiv:2505.20944}
}
read the original abstract

This study explores the effects of Russia's full-scale invasion of Ukraine on the international collaboration of Ukrainian scholars. First and foremost, Ukrainian scholars deserve respect for continuing to publish despite life-threatening conditions, mental strain, shelling and blackouts. In 2022-2023, universities gained more from international collaboration than the NASU. The percentage of internationally co-authored articles remained unchanged for the NASU, while it increased for universities. In 2023, 40.8% of articles published by the NASU and 32,2% of articles published by universities were internationally co-authored. However, these figures are still much lower than in developed countries (60-70%). The citation impact of internationally co-authored articles remained statistically unchanged for the NASU but increased for universities. The highest share of internationally co-authored articles published by the NASU in both periods was in the physical sciences and engineering. However, the citation impact of these articles declined in 2022-2023, nearly erasing their previous citation advantage over university publications. Universities consistently outperformed the NASU in the citation impact of internationally co-authored articles in biomedical and health sciences across both periods. International collaboration can help Ukrainian scholars to go through this difficult time. In turn, they can contribute to the strengthening of Europe.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

7 extracted references · 6 canonical work pages

  1. [1]

    & Di Costa, F

    Abramo, G., D’Angelo, C.A. & Di Costa, F. (2023). Correlating article citedness and journal impact: an empirical investigation by field on a large-scale dataset. Scientometrics, 128, 1877–1894 Abramo, G., D ’Angelo, C. A., & Di Costa, F. (2019). A nation ’s foreign and domestic professors: which have better research performance?(the Italian case). Higher ...

    work page 2023

  2. [133]

    Horizon 2020 Policy Support Facilities (2017). Peer review of the Ukrainian research and innovation system https://projects.research-and- innovation.ec.europa.eu/sites/default/files/rio/report/KI%2520AX%252016%2520008%2520EN%2 520N _UKR_0.pdf Josephson, P. R., & Egorov, I. (1997). The Deceptive Promise o f Reform: Ukrainian Science in Crisis. Minerva, 35(...

    work page 2017

  3. [177]

    OECD (2022). The future of science in Ukraine: Actions now will affect post-war recovery https://www.oecd- ilibrary.org/docserver/afbd05df- en.pdf?expires=1730120081&id=id&accname=guest&checksum=FCE9FEF4AAFB491E641EFEA3 A91 3D8F0 Oleksiyenko, A., Terepyshchyi, S., Gomilko, O., & Svyrydenko, D. (2021). ‘What do you mean, you are a refugee in your own count...

    work page 2022

  4. [261]

    Some issues of the implementation of th e Law on Higher Education https://zakon.rada.gov.ua/laws/show/656-2015-%D0%BF#n17 Cabinet Ministers of Ukraine (2017)

    Cabinet of Ministers of Ukraine (2015). Some issues of the implementation of th e Law on Higher Education https://zakon.rada.gov.ua/laws/show/656-2015-%D0%BF#n17 Cabinet Ministers of Ukraine (2017). On approval of the regulations of state attestatio n of research institutions and higher education institutions https://zakon.rada.gov.ua/laws/show/540-2017-%...

    work page 2015

  5. [521]

    Mammola, S., Fontaneto, D., Martínez, A. et al. Impact of the reference list features on the number of citations. Scientometrics, 126, 785–799 (2021) Martínez, A., & Mammola, S. (2020). Specialized terminology limits the reach of new scientific knowledge. bioRxiv, 2020.08.20.258996. Ministry of Education and Science of Ukraine (2016). On approval of the r...

    work page 2021

  6. [1900]

    Laudel, G

    Journal of the Association for Information Science and Technology, 66(7), 1323-1332. Laudel, G. (2002). Collaboration and reward. Research Evaluation, 11(1) Lawrence, S. (2001). Free online availability substantially increases a paper ’s impact. Nature, 411(6837),

    work page 2002

  7. [1999]

    Mryglod, O., & Mryglod, I

    American Sociological Review, 69(2), 213–238. Mryglod, O., & Mryglod, I. (2020). Collective authorship in Ukrainian scien ce: marginal effect or new phenomenon?. arXiv preprint arXiv:2011.06561. Nogrady, B. (2023). Hyperauthorship: the publishing challenges forbig team 'science. Nature, 615(7950), 175-

    arXiv 2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.