[Paper Review] Gender differences in scientific careers: A large-scale bibliometric analysis

Hanjo D. Boekhout, Inge van der Weijden+1 more|arXiv (Cornell University)|Jun 23, 2021

Sex and Gender in Healthcare31 citations

TL;DR

The paper conducts a longitudinal, cross-country bibliometric analysis of ~6 million researchers from 1996–2018, showing modest but persistent gender differences in publishing careers, productivity, and senior authorship in biomedicine.

ABSTRACT

We present a large-scale bibliometric analysis of gender differences in scientific careers, covering all scientific disciplines and a large number of countries worldwide. We take a longitudinal perspective in which we trace the publication careers of almost six million male and female researchers in the period 1996-2018. Our analysis reveals an increasing trend in the percentage of women starting a career as publishing researcher, from 33% in 2000 to about 40% in recent years. Looking at cohorts of male and female researchers that started their publication career in the same year, we find that women seem to be somewhat less likely to continue their career as publishing researcher than men, but the difference is small. We also observe that men produce on average between 15% and 20% more publications than women. Moreover, in biomedical disciplines, men are about 25% more likely than women to be last author of a publication, suggesting that men tend to have more senior roles than women. Compared with cross-sectional studies, our longitudinal analysis has the advantage of providing a more in-depth understanding of gender imbalances among authors of scientific publications.

Motivation & Objective

• Motivate understanding of gender gaps in scientific careers across disciplines and countries.
• Trace publication careers of nearly six million male and female researchers from 1996 to 2018.
• Quantify how women’s entry into publishing, continued publishing, and senior authorship compare to men.
• Assess longitudinal trends to overcome limitations of cross-sectional studies.
• Provide evidence on how gender differences evolve over time in global science.

Proposed method

• Perform a large-scale bibliometric analysis across all scientific disciplines and many countries.
• Adopt a longitudinal design to track publication careers from 1996–2018.
• Compare cohorts by start year to analyze continuation and productivity differences.
• Measure metrics such as publication counts and last-author status in biomedical fields.
• Contrast longitudinal findings with cross-sectional studies to assess imbalances.
• Quantify gender differences in senior authorship in biomedical disciplines.

Experimental results

Research questions

• RQ1 What are the gender differences in starting a publishing career over time on a global scale?
• RQ2 Do women continue publishing at similar rates as men when accounting for cohort effects?
• RQ3 How does average productivity (publications) differ by gender, and by discipline?
• RQ4 Is there a gender gap in senior (last) authorship, particularly in biomedical fields?
• RQ5 How do longitudinal patterns compare to cross-sectional findings in explaining gender imbalances?

Key findings

• The share of women starting a publishing career rose from 33% in 2000 to about 40% in recent years.
• Women appear somewhat less likely than men to continue their publishing careers, but the difference is small.
• Men publish on average between 15% and 20% more publications than women.
• In biomedical disciplines, men are about 25% more likely than women to be last author.
• Longitudinal analysis provides a deeper understanding of gender imbalances than cross-sectional studies.