The division of labor within a research paper plays a crucial role in fostering efficient collaboration and knowledge innovation. The authors' engagement in different contributions influences the integration of specialized knowledge, the formation of diverse perspectives, and the stimulation of creativity, which in turn impact the novelty level of the paper. However, previous studies have lacked depth in exploring the relationship between paper division of labor and novelty, and have overlooked potential gender differences. This study, based on 81,137 papers from PLOS ONE, investigates the correlation between authors' contributions engagement, contributions engagement of authors of different genders, and paper novelty. The results show that, in the Writing-original draft preparation, Writing-review & editing, Methodology, and Software, a higher proportion of author participation is associated with a higher likelihood of the paper achieving greater novelty. In terms of gender differences, women are more likely to participate in the Investigation, Data curation, Formal analysis, and Writing-original draft preparation, while men tend to play a more prominent role in Supervision, Resources, Funding acquisition, Conceptualization, and Software. Furthermore, the study shows that, regardless of gender, a greater proportion of participation in the Writing-original draft preparation, Writing-review & editing, and Software is significantly associated with higher paper novelty. However, only for male authors, a greater proportion of participation in Methodology, Visualization, and Funding acquisition is associated with higher paper novelty.
Scientific collaboration is defined as “a concerted effort by researchers to achieve a common goal of
generating new scientific knowledge” [
RQ1: Is author's participation in different research contributions in papers correlated with paper novelty?
Previous research on author division of labor based on author contribution statements has
revealed significant differences in division of labor among scholars of different genders. Haeussler
et al., based on data from 12,964 papers from PLOS ONE, found that women are more likely to
participate in experimental rather than conceptual activities [
RQ2: Is the research contribution engagement of authors of different genders correlated with paper novelty?
PLOS ONE is a multidisciplinary open access journal that supports the development of knowledge
dissemination. Therefore, this study selects articles published in PLOS ONE as the research data,
retrieving a total of 124,688 papers published between 2016 and 2024 from the PLOS ONE journal
website0. Focusing on gender differences in the division within collaborations, we exclude 1,860
single-authored papers. Author gender identification is conducted using the Genderize.io tool0, and
papers with incomplete gender identification are removed (11,985 papers), leaving 110,843 papers
for analysis. Lin et al. developed the SciSciNet dataset, which encompasses over 134 million
scientific publications and millions of external links related to funding and public uses, providing
metrics such as paper novelty [10]. We use the paper DOIs to match PLOS ONE papers with
SciSciNet records, obtaining the tail novelty metric (Atyp_10pct_Z), developed by Uzzi et al. [
Since 2016, PLOS ONE has adopted the CRediT contribution taxonomy, encompassing 14 research contributions 0. Based on 81,137 PLOS ONE papers, this study investigates the relationship between the author participation rate in each contribution (as shown in Formula 1), the male author participation rate (as shown in Formula 2), the female author participation rate (as shown in Formula 3), in relation to the novelty of the paper (Atyp_10pct_Z). The regression model controls for team size (Teamsize), the proportion of contribution categories involved in each paper (All_Contribution_per, as shown in Formula 4), the publication year (Fixed year), the proportion of female/male authors in the paper (Per_f/Per_m).
P_i=Contribution_i_autho rs (1)
Total_authors P_m_i=Contribution_i_male_authors (2)
Total_authors P_f_i=Contribution_i_female_authors (3)
Total_authors All_Contribution_per=CRediT_contributions (4) 14 Where i refers to one of the 14 contributions defined in the CRediT taxonomy: 'Conceptualization', 'Data curation', 'Formal analysis', 'Funding acquisition', 'Investigation', 'Methodology', 'Project Administration', 'Resources', 'Software', 'Supervision', 'Validation', 'Visualization', 'Writing-original 0 https://journals.plos.org/plosone/ 0 https://genderize.io/ 0 https://credit.niso.org/ draft preparation', and 'Writing-review & editing'. Contribution_i_authors represents the number of authors contributing to i per paper. Total_authors represents the total number of authors per paper. Contribution_i_male_authors and Contribution_i_female_authors signify the respective counts of male and female authors contributing to i per paper, and CRediT_contributions represent the contribution categories of CRediT involved in each paper. 3. Results This study investigates whether gender-based preferences exist in the division of labor among authors of PLOS ONE publications. It explores the relationship between author contribution engagement and paper novelty, specifically examining whether this relationship differs between male and female authors.
The 81,137 papers from PLOS ONE involve a total of 534,898 authors, with 208,733 female authors (39%) and 326,165 male authors (61%). Figure 1 presents the author participation rates in contributions and the difference in participation rates between genders. In Figure 1, All represents the proportion of authors participating in i to the total number of authors in all papers. F-M represents the difference in participation rate between male and female authors. It is calculated as the proportion of female participation in i (i.e., the total number of female authors performing i in all papers / the total number of female authors) minus the proportion of male participation in i (i.e., the total number of male authors performing i in all papers / the total number of male authors). Here, i belongs to the 14 contributions of CRediT.
All
F-M Resources, Data curation, Investigation, Validation, and Supervision are associated with lower novelty.
To address RQ2, two additional multiple linear regression analyses (Models 2 & 3 in Table 1) are conducted, focusing specifically on the relationship between the participation rates of female and male authors in various contribution roles and paper novelty. The regression results, presented in Table 1, reveal that a greater proportion of female author participation in Writing-original draft preparation, Writing-review & editing, and Software contributions is significantly associated with higher paper novelty. Conversely, a greater proportion of female author participation in Resources, Data curation, Formal analysis, and Investigation is associated with lower paper novelty. For male authors, a greater proportion of participation in Visualization, Writing-original draft preparation, Writing-review & editing, Methodology, Software, and Funding acquisition is significantly associated with higher paper novelty. However, a greater proportion of male author participation in Resources, Data curation, Investigation, Validation, and Supervision is associated with lower paper novelty.
The findings of this section not only confirm the correlation between author participation rates in different contribution roles and paper novelty, but also reveal the differences in the relationship between the participation rates of authors of different genders and paper novelty. Comparing the relationship between male and female author participation rates and paper novelty, it is found that only for males, participation in Visualization, Methodology, and Funding acquisition contributions is positively correlated with paper novelty. This finding provides a new research perspective for exploring the potential relationship between author gender and innovative outputs in scientific collaboration.
This study investigates the relationship between gender differences in author division of labor and paper novelty. Our findings indicate that women are more likely than men to participate in contribution roles related to analysis, data curation, and writing. Men, on the other hand, are more likely to participate in contribution roles involving Supervision, Funding acquisition, and Conceptualization of the paper. The results show that papers with a higher proportion of authors
VARIABLES Per_m Teamsize All_Contribution_per P_m_conceptualization P_m_resources P_m_visualization P_m_ writing-original draft preparation P_m_writing-review & editing P_m_data curation P_m_formal analysis P_m_investigation P_m_methodology P_m_software P_m_validation P_m_funding acquisition P_m_project administration P_m_supervision Fixed year Constant Observations R-squared participating in the Writing-original draft preparation, Writing-review & editing, Methodology, and Software contributions are more likely to exhibit higher novelty. A greater proportion of both male and female authors’ participation in Writing-original draft preparation, Writing-review & editing, and Software contributions is significantly associated with higher paper novelty. Additionally, for male authors, a greater proportion of participation in Visualization, Methodology, and Funding acquisition is significantly associated with higher paper novelty. The findings of this study provide reference suggestions for optimizing the division of labor in research teams to achieve the production of highly novel papers.
This study’s data is confined to PLOS ONE publications. Future research could extend this
analysis to other academic journals, investigating the relationship between author contribution
engagement and paper novelty across different publication venues. Additionally, this study
employs the Atyp_10pct_Z metric, developed by Uzzi et al. [
The paper is presented at the second Workshop on “Innovation Measurement for Scientific Communication (IMSC) in the Era of Big Data” at 2024 ACM/IEEE Joint Conference on Digital Libraries (JCDL). This work was partially supported by the National Natural Science Foundation of China (No. 72074113).
During the preparation of this work, the authors used GPT-4 in order to correct grammatical errors, typos, and other writing mistakes. After using this tool, the authors reviewed and edited the content as needed and takes full responsibility for the publication’s content. [10] Z. Lin, Y. Yin, L. Liu, D. Wang, SciSciNet: A large-scale open data lake for the science of science research, Scientific Data, 10 (2023) 315.