Source-linked AI summary
Gender differences in research collaboration
Giovanni Abramo, Ciriaco Andrea D'Angelo, Gianluca Murgia
TL;DR
Gender differences in research collaboration remain less studied than other dimensions of academic gender inequality. Using individual-level bibliometrics, the paper compares collaboration propensities across fields, disciplines, and collaboration forms among Italian academics. Women show greater propensity for general, intramural, and domestic extramural collaboration, while men retain an international-collaboration advantage.
Problem
Gender differences in research collaboration have received limited attention despite the increasing collaborative character of scientific research.
Method
The study calculates collaboration propensity for each scientist from publication records and compares gender averages across fields, disciplines, and collaboration forms.
Results
Women show greater propensity for general, intramural, and domestic extramural collaboration, while men show greater propensity for international collaboration overall and in most fields.
Takeaways & Limitations
Evaluations using international-collaboration indicators require caution because they may penalize women academics and organizations with more women.
Takeaways & Limitations
The analysis is limited to university researchers because academic identities and disciplines cannot be assigned unequivocally for non-academic authors.
Abstract
from arXiv · showhide
The debate on the role of women in the academic world has focused on various phenomena that could be at the root of the gender gap seen in many nations. However, in spite of the ever more collaborative character of scientific research, the issue of gender aspects in research collaborations has been treated in a marginal manner. In this article we apply an innovative bibliometric approach based on the propensity for collaboration by individual academics, which permits measurement of gender differences in the propensity to collaborate by fields, disciplines and forms of collaboration: intramural, extramural domestic and international. The analysis of the scientific production of Italian academics shows that women researchers register a greater capacity to collaborate in all the forms analyzed, with the exception of international collaboration, where there is still a gap in comparison to male colleagues.
1. Introduction
Research collaboration has received less attention in gender-gap research despite its growing role in scientific production. This study addresses that gap with an individual-level bibliometric analysis.
- Gender research has focused mainly on male overrepresentation, productivity differences, and discrimination.
- Collaboration is increasingly central to scientific research and is positively correlated with scientific productivity.
- Earlier collaboration studies primarily used surveys to investigate why male and female researchers behave differently.
- Bibliometric studies of collaborations initiated by women and men remain rare and limited to restricted contexts.
- The study advances this literature through large-scale bibliometric analysis at the individual level.
2. Gender aspects in research collaboration: literature review
Prior research reports conflicting gender patterns in collaboration and identifies social, institutional, and family-related constraints on women’s networks. This study broadens the comparison across collaboration forms, disciplines, and the full population of Italian academics examined.
- Earlier studies variously report that women collaborate differently, less effectively, or with less cosmopolitan and prestigious networks than men.
- Collaboration may have a particularly strong productivity impact for women by helping them overcome limited social capital and integrate into academia.
- Gender homophily can direct women toward same-gender collaborators, increasing isolation because women remain minorities in principal disciplines.
- Women’s academic isolation is associated with masculine cultures, smaller departments, and disciplines dominated by established male networks.
- International collaboration may be constrained by family ties, prejudice in some countries, and conflicting evidence about communication technologies.
- The study examines general, intramural, domestic, and international collaboration across 11 fields and 200 disciplinary levels using individual scientists as the analytical unit.
3. Methodology, dataset and indicators
The study measures collaboration propensities from individual academics’ publication records in a large dataset of Italian researchers. It compares women and men across collaboration forms, fields, and disciplines while addressing distortions from aggregate bibliometric measures.
- Aggregate gender-level co-authorship ratios can be distorted because scientific production is highly skewed across academics.
- The analysis instead calculates each scientist’s co-authored-publication ratio and averages these individual propensities by gender.
- The dataset covers Italian academics in bibliometrically observable hard-science and selected social-science fields, excluding arts and humanities.
- The dataset identifies academic authors, genders, disciplines, universities, coauthors, and domestic or foreign institutional addresses.
- The field of observation includes 11 UDAs and 200 SDSs meeting the requirement that at least half of academics produced a publication during 2006–2010.
- The four indicators measure general, intramural, domestic extramural, and international extramural collaboration propensities on a scale from 0 to 1.
- The dataset reports 36,211 productive academics and 197,460 publications, while female and male collaborative academics comprise 83.7% and 82.0%, respectively.
4. Results and discussion
The study compares gender differences in collaboration propensity across Italian academic fields and disciplines, distinguishing intramural, domestic extramural, and international extramural collaboration. Women generally show higher collaboration propensity, but international collaboration remains an exception, with discipline-level variation.
- Measurement and scope: The analysis compares female and male academics using collaboration-propensity indicators C, CI, CED, and CEF across different UDAs.The indicators cover general, intramural, domestic extramural, and international extramural collaboration; Mann-Whitney tests assess gender differences.
- Measurement and scope: Single-authored papers with multiple affiliations are excluded, while publications combining collaboration forms contribute to each form observed.This operational rule allows one publication to count toward both intramural and extramural collaboration when applicable.
- Overall collaboration: Women generally have higher overall collaboration propensity and a higher percentage of co-authored publications, although exceptions occur in selected disciplines.Women are below men in the percentage of collaborative academics in Physics, Mathematics and computer sciences, and Pedagogy and psychology, yet their co-authored-publication share remains higher.
- Intramural collaboration: Women show greater intramural collaboration propensity overall and in most UDAs, with Civil engineering the only reported exception and not statistically significant.In Civil engineering, women have higher shares with nil intramural propensity and lower shares with completely intramural publications.
- Domestic extramural collaboration: Women show greater domestic extramural collaboration propensity overall, but the pattern varies substantially across UDAs.The female advantage is confirmed in Physics, Medicine, Earth sciences, and Industrial and information engineering, with statistical significance reported only for Industrial and information engineering.
- International extramural collaboration: Men show greater international extramural collaboration propensity overall and in nearly every UDA, with Earth sciences the only female-led exception.Many reported Mann-Whitney differences are not statistically significant.
- Cross-form comparisons: Earth sciences is the only UDA where women exceed men across all collaboration forms, while Civil Engineering shows the opposite pattern across sub-forms.In several other UDAs, women exceed men generally and intramurally but trail them in domestic and international extramural collaboration.
- Relationships among collaboration forms: Correlations among collaboration-propensity indicators are generally strongly significant but often below 0.40, with relationships differing by UDA and gender.C is positively correlated with CI in only certain UDAs, CI and CED are negatively correlated, and CED and CEF are strongly positively correlated only in Physics.
5. Conclusions
Using an individual-level bibliometric approach, the study finds that Italian women academics generally collaborate more than men, except internationally, with differences linked to discipline.
- The study addresses limited research on gender differences across the possible forms of research collaboration.
- The analysis measures each academic’s propensity to initiate collaboration, reducing distortion from highly collaborative outliers.
- Female researchers show greater propensity to collaborate overall, intramurally, and with domestic institutions than male researchers.
- International collaboration remains the principal exception, with women showing a lower propensity than men.
- Collaboration gaps and relationships among collaboration forms vary according to academics’ discipline of specialization.