Mapping Gender in Research: the case of Portugal

Authors

Elsa Fontainha, Tanya Araújo

Lisbon School of Economics and Management (ISEG) Universidade de Lisboa

Contact: elmano@iseg.ulisboa.pt

Abstract

The research “Mapping Gender in Research: the case of Portugal” is the Case Study developed by ISEG under the PLOTINA project.  The goal of this Case Study is to map, by gender, the research in all scientific domains in Portugal during the last decade. The main databases and sources are: Web of Science, WoS; Portuguese Science Foundation (Fundação para a Ciência e Tecnologia FCT); Quadros de Pessoal (GEP-MTSSS); European Union Research Projects (CORDIS – EU research projects under FP7 (2007-2013) and EU H2020 Projects). Focusing on gender, the research will make possible to characterize in detail and for the first time in Portugal the research done by gender (e.g. by RPOs, scientific areas, quality, dynamics, collaborative work). A new database will be built and made available for the research community.

Aims: These case studies contribute to:

  • to identify the real contribution of women in science and put an end to some myths about it;
  • to increase the visibility of the research made by women incentivizing the attractiveness of new generations of women for research;
  • to show the collaborative patterns and networks established by women in the scientific communities in the different fields of research and among countries;
  • to provide a new open dataset to be used in the future by all the scientific community in particular in Europe;
  • to provide Research Performing Organizations (RPOs) in Portugal and abroad of a new information source;
  • call the attention for the lack of information about gender in some crucial domains and the urgent necessity to solve that lack of information.

Methodologies

To build the database

In most of these sources of data the detailed information by women and men is not available directly or indirectly. For example, the bibliographic database Web of Science Data (WoS), includes for the period 2006-2016, a total of 108,882 articles in English published by researchers affiliated to Portuguese Research Performing Organisations (RPOs) but no information of authors’ gender. In order to accurately allocate the gender of authors a self-tailored algorithm was built and applied. The methodology of gender authorship disambiguation is based on first name. This methodology is applied by Larivière et al. (2013) and many other authors and referred and discussed to in Araújo & Fontainha (2017 a), b), c) and Fontainha & Araújo 2018). Focusing on gender, the case study with the big database created, contributes to characterize in detail and for the first time in Portugal the research done (e.g. by RPOs, scientific areas, quality, dynamics, collaborative work).

To analyse the results

The methodology to apply to the data includes, among other: descriptive statistics, complexity analysis and network analysis. Some examples are provided in next section: Gendered Results.

 

Gendered Results: The main results, with some additional information in next point (Full Case Study) are:

-The proportion of top positions in the academic staff (European Commission, 2016 and 2019) contrasts with the scientific performance evaluated by research outputs resulting from the Mapping Gender in Research: the case of Portugal.

-The increase of scientific research in Portugal corresponds to an increase of the women participation as main (correspondent/principal) authors of the papers.

-The patterns of interdisciplinary collaboration for men and women are different.

-The international collaboration networks are different between men and women and are denser for men than for women.

-The singular authorship follows different standards for researchers.

Introduction

The study of the Portuguese scientific production by gender is particularly relevant because:  (i) Women have higher level of education  in many fields (European Commission, 2016 and 2019); (ii) The proportion of top positions in the academic staff contrasts with the scientific performance (European Commission, 2016 and 2019); (iii) The proportion of women working as researchers (the distribution of R&D personnel across occupations in the higher education sector) is particularly high in Portugal (European Commission, 2019). (iv)Portugal, compared with 11 other countries/regions, has the highest share of women as inventors (26%) and in the research population (49%) (Elsevier, 2017) and in 20 of 27 scientific subjects  has the highest share of women among researchers (Elsevier, 2017, p. 22).

The case study includes the creation of a database open to scientific community in order to improve the study of women in research, teaching and training. The following topics illustrate some possibilities of research based on that database: Authorship by Gender, International Research Collaboration by Gender, Interdisciplinary Research Collaboration by Gender.

Authorship by Gender

After downloading the original information from Web of Science- Clarivate Analytics (WoS©) the gender authorship was allocated based on a self-tailored algorithm developed in MathLab software. The scientific production of the researchers affiliated in Portuguese RPO shows a sharp increase in the last years and the women are the most frequent reference author (main/correspondent author) since 2011 (Figure 1). In nine years the ratio woman/men moved from 0.9 to 1.1. The Figure 1 represents a total of 23,981 articles published in Social Sciences and also indicates the number of articles n.a., it means, not allocated by gender. This lack of observations is a problem that exists in all studies using the same methodology. However, the percentage of not allocated is 15% in 2016 and there is a clear trend for decrease. That percentage of not allocated decreases for 4% when the articles analysed are only those single authored (Figure 2). The articles with only one author represent a decreasing share of the total of articles (from 22% in 2008 to 17% in 2016) and there is some differentiation between men and women since 2014 however more years must be included in the analysis to conclude about the preference of male researcher to publish alone.

International Research Collaboration by Gender

The international collaboration in research is studied based on the different country allocation of the co-authors. Because each author can have different affiliations in different countries the unique pair author-country was done following two criteria: if the author has an affiliation in a Portuguese RPO the affiliation is allocated to Portugal; and if the author has several country allocations the one considered is the first/main indicated in the article. An example of the results in 2016 for the networks of international research collaboration when there are at least researchers from three different countries are presented in Figure 3. The international collaboration networks are different between men and women and are denser for men than for women. Concerning the countries, the largest difference between men and women is the high level of collaboration of men with United States and United Kingdom, where there are most of the excellence RPO. For women, the language proximity seems to have a role in international research collabolation, having Spain and Italy a relevant position (Figure 3).

Interdisciplinary Research Collaboration by Gender

The interdisciplinary results are illustrated with the scientific domain of Economics, for the period 2010-2015 and whose authors are affiliated to a Portuguese institution. Details about this study are available (open source) in Araújo & Fontainha (2017a). The methodology adopted is to organize the articles by categories of scientific domains and by categories of authorship and apply descriptive analysis and after apply the measure Minimum Spanning Tree (MST) to the network created. The main results are:

-The woman inclusive articles (it means articles authors that include at least one woman) is the category of articles with the second highest average number of authors. These results converge to the hypothesis that women prefer to work in teams. However, when papers are exclusively authored by women, the working teams tend to be smaller than any of those that also include men.

-The interdisciplinary publications (and the consequent cross-fertilization of scientific domains) between the research of economists and other scientists are also different by gender. Academic women compared with their male counterparts reveal preference for the subjects Environmental Sciences, Management and Political Sciences. On the other hand, the subjects Social Sciences, Mathematics and Finance display higher frequencies in papers either inclusively (it means articles authors that include at least one man) or exclusively authored by men.

– Gender differences in collaborative research and interdisciplinarity in scientific outputs have received little attention when compared with the growing importance that women hold in academia and research.

The adoption of a network approach allowing to uncover the emergence of a specific pattern when the network of scientific subjects is induced from a set of papers exclusively authored by men. Such a male exclusive authorship condition is found to be the solely responsible for the emergence of that specific shape in the structure of the network (Figures 4 and 5).

Fact and figures

Figure 1: Articles in Social Science by Gender, Portugal (2008-2016) (N=23,981)

Figure 2: Single Authorship by Gender Portugal (2007-2016) (N=4,039)

Figure 3 – International Research Collaboration by Gender in Social Sciences (2016)

Figure 4: The Minimum Spanning Tree of women exclusive category

Figure 5:  The Minimum Spanning Tree of men exclusive category

References

Araújo, Tanya & Fontainha, Elsa (2016). The specific shapes of gender imbalance in scientific authorships: a network approach, Working Papers Department of Economics 2016/17, ISEG – Lisbon School of Economics and Management, Department of Economics, Universidade de Lisboa. LINK (open source)https://depeco.iseg.ulisboa.pt/wp/wp172016.pdf

Araújo, Tanya & Fontainha, Elsa (2017a). The specific shapes of gender imbalance in scientific authorships: a network approach, arXiv preprint arXiv: 1608.07224v2. LINK (open source): https://arxiv.org/pdf/1608.07224v2.pdf

Araújo,  Tanya, & Fontainha, Elsa (2017b). Big Missing Data: are scientific memes inherited differently from gendered authorship?. arXiv preprint arXiv:1706.05156. LINK (open source): https://arxiv.org/pdf/1706.05156.pdf

Araújo, Tanya & Fontainha, Elsa (2017c). The specific shapes of gender imbalance in scientific authorships: a network approach. Journal of Informetrics, 11(1), 88-102.

Elsevier (2017). Gender in the global research landscape—analysis of research performance through a gender lens across 20 years, 12 geographies, and 27 subject areas. Elsevier.

European Commission (2016) . She Figures 2015.

European Commission (2019) . She Figures 2018.

Fontainha, Elsa & Araújo, Tanya (2018). Mapping Gender in Research: the case of Portugal- Case Study PLOTINA Project.In Proceedings of 3rd Forum of Research, Research in Social Sciences & Management CSG, Lisbon, 24 May 2018. LINK (open source): https://csg.rc.iseg.ulisboa.pt/wp-content/uploads/2019/01/artigo11_Mapping-gender-in-research.pdf

Fontainha, Elsa & Araújo, Tanya (2019, forthcoming). Mapping gender in research: the international collaboration networks. Case Study PLOTINA PROJECT. In Proceedings of 4th Forum of Research, Research in Social Sciences & Management CSG, Lisbon, 9 May 2019. LINK (open source): https://csg.rc.iseg.ulisboa.pt/ (forthcoming)…

 

Larivière, V., Ni, C., Gingras, Y., Cronin, B., & Sugimoto, C. R. (2013). Global gender disparities in science. Nature, 504(7479), 211–213.