Source-linked AI summary
Gender differences in research productivity: a bibliometric analysis of the Italian academic system
Giovanni Abramo, Ciriaco Andrea D'Angelo, Alessandro Caprasecca
TL;DR
The paper addresses persistent evidence of gender differences in research productivity by examining the Italian academic research system. Using bibliometric data for the full population of researchers in selected scientific-technological disciplines, it finds significant but generally smaller differences, with sectorial variation and some areas where women are not inferior.
Problem
Prior research generally reports better research productivity for men, while evidence for the Italian academic research system and variation across sectors and roles remains limited.
Method
The study conducts a bibliometric analysis of the entire technological-scientific population of Italian university researchers, comparing gender and academic roles across multiple performance indicators.
Results
Gender differences are significant but smaller than much of the prior literature suggests, vary by scientific sector, and are more apparent for quantitative indicators than other measures.
Takeaways & Limitations
Men generally perform better, but women are not inferior in a significant number of scientific sectors.
Takeaways & Limitations
The study does not account for researchers’ time devoted to research, including maternity leave, which may amplify observed differences for assistant professors.
Abstract
from arXiv · showhide
The literature dedicated to analysis of the difference in research productivity between the sexes tends to agree in indicating better performance for men. This study enters in the vein of work on the subject. Through bibliometric examination of the entire population of research personnel working in the scientific-technological disciplines of Italian university system, it confirms the presence of significant differences in productivity between men and women. But such differences result as being smaller than reported in a large part of the literature, confirming an ongoing tendency towards decline, and are also seen as more noticeable for quantitative performance indicators than other indicators. The gap between the sexes presents important sectorial specificities. In spite of the generally better performance of men, it can not be ignored that there are a significant number of scientific sectors in which the performance of women does not result as inferior.
1. Introduction
The study examines gender differences in scientific research productivity, extending prior findings to the entire technological-scientific population of Italian universities. It evaluates whether differences vary across sectors and academic roles while leaving their causes for future investigation.
- Prior research generally finds lower publication rates for women across countries, disciplines, periods, and measures.
- Evidence also indicates that gender differences in scientific performance lessen over time and partly reflect specialization or academic position.
- The study provides evidence from the Italian academic research system for the first time.
- It asks whether productivity differences exist, whether they appear across evaluation parameters, and whether they show sectorial or employment-level specificities.
- The study does not investigate the causes of observed gender differences, but identifies directions for further research.
- The analysis covers approximately 33,000 Italian university research scientists and classifies each by academic role and scientific field using publications from 2001 to 2003.
2. Factors affecting scientific production
The paper frames research productivity as the outcome of personal and environmental factors that must be considered when comparing scientists. It therefore accounts especially for gender, academic role, specialization, and institutional context.
- Research outputs include tangible products such as publications, patents, and conference presentations, alongside intangible outputs such as knowledge and consulting.
- Comparisons between researchers should filter factors other than individual merit that may affect performance directly or indirectly.
- The literature identifies gender, age, marriage, children, specialization, academic rank, teaching load, and institutional prestige as productivity determinants.
- Academic rank is associated with productivity, with studies reporting higher publication rates for full professors than for associate professors and research staff.
- Men generally perform better in research, whereas women may favor teaching activities more than men.
- The study considers relevant personal and environmental factors while using a broader and more representative observation field than preceding studies.
3. Analytical model
The analytical model compares Italian university researchers by gender and academic role using bibliometric indicators derived from SCI-indexed publications. It incorporates disciplinary and institutional considerations while documenting important measurement and coverage limitations.
- The analysis focuses on gender and academic role, with specialization implicitly controlled through scientific-disciplinary sectors.
- The dataset covers eight scientific-technological areas and excludes civil engineering and architecture because SCI is not sufficiently exhaustive for that research output.
- Publication data come from the Italian Observatory of Public Research, which extracts SCI literature and disambiguates publication origins.
- Researchers lacking a staff role throughout 2001–2003 or changing scientific-disciplinary sector were removed from the dataset.
- Performance indicators include output, fractional output, contribution intensity, scientific strength, fractional scientific strength, and quality index.
- Journal publications dominate submitted research products, exceeding 90% in seven of the eight discipline areas considered.
- Journal impact factors are normalized by disciplinary-sector averages, but journal quality remains an imperfect proxy for publication quality.
- The model cannot quantify researchers’ time devoted to research or maternity and sick leave during the observation period.
4. Results
Across Italian academic research, men generally show higher average performance, but gender differences vary by indicator, professional role, and disciplinary sector. Women match or exceed men in substantial subsets of sectors, while male output is more concentrated among highly productive researchers.
- 61.5% of Italian academic research personnel published at least one SCI-listed article during the triennium, with nearly identical inactivity rates for women and men overall.Women’s inactivity rate was 38.6%, compared with 38.5% among men; within roles, women were more active than men by 1.1% among full professors, 1.5% among associates, and 5.2% among assistants.
- 38% of active scientists produced less than one article annually, while the top 20% generated over 53% of total scientific production.Publication frequency was skewed, with women’s distribution less concentrated than men’s; skewness was 2.25 for women and 3.46 for men.
- 4.1 Differences in performance: Men had higher concentration indexes than women across all professional roles, consistent with a larger concentration of highly productive or “star” scientists among men.The indexes comprised the cumulative production of the first two deciles and the Gini coefficient.
- 4.1 Differences in performance: Men’s average output exceeded women’s by 13.3% among full professors, 12.3% among associates, and 17.5% among assistants.For scientific strength, the corresponding male advantages were 19.7%, 15.9%, and 20.2%.
- 4.1 Differences in performance: The output percentile gap narrowed with professional seniority, from 4.8% among assistants to 3% among associates and 1.5% among full professors.The authors note that assistant-professor differences may be amplified because maternity leaves were not accounted for.
- 4.2 Analysis at the level of single sectors: Women were not inferior to men in output in 43 of 110 full-professor sectors, with substantial sector-specific reversals across engineering, agriculture, veterinary, earth, medical, and related sciences.Within-sector rankings showed no correlation between the number of women and their ranking relative to men.
5. Final considerations
Using the entire technological-scientific population of the Italian academic system, the study confirms higher average male performance but finds a smaller, mainly quantitative gap than much prior literature. It also identifies substantial sectorial variation and many sectors where women are not inferior.
- Men showed higher average productivity than women across all evaluated performance indicators, with the largest gap in quantitative output.
- The average gender gap was smaller than results reported by other authors, although the study cannot make inter-temporal comparisons.
- Women were not inferior in output in 43 of 100 sectors for full professors, 55 of 146 for associates, and 49 of 147 for assistants.
- Women were more concentrated at the lowest productivity levels, whereas men were more concentrated at the highest levels, possibly reflecting more male “star scientists.”The authors identify this as a suggestion for future work rather than a demonstrated explanation.