=Paper= {{Paper |id=Vol-2709/paper222 |storemode=property |title=Ecuadorian Scholarly Production in Computer Science. Analysis of Publication Patterns by Gender |pdfUrl=https://ceur-ws.org/Vol-2709/paper222.pdf |volume=Vol-2709 |authors=Janneth Chicaiza,Ruth Reátegui |dblpUrl=https://dblp.org/rec/conf/clei/ChicaizaR20 }} ==Ecuadorian Scholarly Production in Computer Science. Analysis of Publication Patterns by Gender== https://ceur-ws.org/Vol-2709/paper222.pdf

Ecuadorian Scholarly Production in Computer
Science. Analysis of Publication Patterns by
Gender

Janneth Chicaiza1[0000−0003−3439−3618] and Ruth Reátegui1[0000−0002−7304−4413]

Computer Science Department
Universidad Técnica Particular de Loja, Ecuador
jachicaiza@utpl.edu.ec, rmreategui@utpl.edu.ec

Abstract. Men and women compose the human capital and workforce
needed for the economic and social development of a country. Despite
this, in science, there is a gender participation gap, unfavorable for
women. This statement is supported by a large body of literature that
describes gender diﬀerences in scholarly productivity. But, mainly in the
Computer Science field, there is little evidence about publishing pat-
terns by gender in countries with a recent path in scholarly publishing.
To meet the information gap of Ecuador, in this study, the authors focus
on analyze publishing patterns by gender in Computer Science. Specifi-
cally, this paper examines (1) gender diﬀerences in research productivity
and participation rates, and (2) publishing patterns regarding document
type and impact. Findings show that despite the great diﬀerence in the
number of female and male researchers that leading scientific production,
both have common publishing patterns. Except, when a woman is the
first author of a document, co-authors tend to distribute equitably by
gender. Whereas when a man leads the research, the presence of women
is reduced.

Keywords: Scholarly production · gender analysis · publishing patterns
· Ecuador · Scopus

1 Introduction

Men and women compose the human capital and workforce needed for the eco-
nomic and social development of a country. Likewise, in sciences, both female and
male researchers are essential to creating and disseminating knowledge. Despite
this, there is a gender participation gap unfavorable for women. More precisely,
in STEM (Science, Technology, Engineering, and Mathematics) fields, women
perform research activities in fewer numbers than men.
In scholarly production, an increase in female authors has been observed,
from 12% in 1995 to 35% in 2005 [6]. But, there are diﬀerences with respect
c 2020 for this paper by its authors. Use permitted under Creative Com-
Copyright ©
mons License Attribution 4.0 International (CC BY 4.0).
to the several disciplines in which women work, being 15% of participation in
mathematics, physics and computer science. According to UNESCO Institute
for Statistics (UIS), less than 30% of women in the world are researchers and
numerous studies have shown that women in STEM fields publish less, their
salaries are lower, and they do not have the same progress as their peers.
In Ecuador since the Constitution of 2008, equality has been promulgated for
anyone who is in a situation of inequality. In the Magna Carta of 2008, access
to education has been established as well as the fundamental rights of women
living in Ecuador. Despite this, in 2014, 44.6% of full-time R&D personnel are
women in Ecuador [14], which confirms that there is also a gender imbalance in
terms of research.
There is a large body of literature describing gender diﬀerences in scholarly
publishing productivity. Diﬀerences in productivity between men and women
employed in the scientific world have always attracted interest from a wide range
of observers [13,9]. Causes and eﬀects [1,8] of gender imbalances in publishing
have been studied widely in diﬀerent contexts, in diﬀerent countries, and in
diﬀerent scientific fields [12].
For example, [3] analyze whether two German programs (German Women
Professorship Program and Pact for Research and Innovation) have allowed in-
creasing the participation of women in leadership positions, as well as increasing
their scientific publications. Also, authors of [4] analyze characteristics of women
publications in the areas of physics and astronomy in India. Considering the same
country, [5] presents an analysis of the publications made by Indian scientists in
disciplines related to life science. The work of [6] presents an analysis of gender
diﬀerences in the scientific career considering the information from 83 countries
in 13 disciplines. Another work with a diﬀerent approach in terms of analysis
techniques is carried out by [2] who used network or graph theory to identify
the importance of gender in patterns of scientific collaboration.
Furthermore, in [10], authors analyze the contribution of Iranian women to
scientific production, emphasizing the characteristics of women’s publications
rather than a comparison with men. In the work of [7] they carried out an
analysis of the factors that inhibit the research productivity of Iranian women,
collecting the information with a questionnaire.
Most of these studies have focused on the analysis of diﬀerent STEM carriers.
Regarding the Computer Science (CS) field, [15] studied the participation of
women as authors and editors in publications related to software engineering.
This work includes information of authors from 61 diﬀerent countries. So far,
there are few works that analyze the contribution of female researchers from
STEM careers in countries of South America with a recent path in scholarly
publishing. Mainly in the CS field, there is little evidence about publication
patterns by gender in countries with a recent path in scientific publishing. To
meet the information gap, in this paper, authors try to provide evidence about
publishing patterns from the Ecuadorian researchers. Specifically, this paper
examines (1) gender diﬀerences in research productivity in Computer Science,
and (2) publishing patterns between women and men regarding document type
and impact.

2 Methodology
In this section, the authors explain the two main stages that they have carried out
to analyze Ecuadorian production in Computer Science by researchers’ gender:
data gathering and cleaning, and scope definition and data analysis.

2.1 Data gathering and cleaning
Production data of Ecuadorian researchers were collected from Scopus. This
multidisciplinary database was chosen because it is the largest abstract and
citation source of peer-reviewed literature. Scopus was launched by Elsevier in
2004 and indexes documents published in more than 22000 active sources. Scopus
in addition to indexing abstracts and citations, allows authors and institutions
to create profiles to facilitate the organization and searching for resources using
diﬀerent criteria.
To access Scopus information, people use search functionalities provided by
the database through the user interface (UI). In addition to the UI, Scopus pro-
vides access from a set of Application Programming Interfaces (APIs), which
oﬀer the same functionality but in a machine-readable format. To obtain in-
formation about the publications of authors with aﬃliation from Ecuador we
use the Scopus APIs. Diﬀerent methods provide metadata about publications,
authors, institutions, sources, and more.
To get data of Computer Science documents, published until 2018, in which
at least one author with an Ecuadorian aﬃliation has participated, the following
query was sent to the Scopus Search API.
AFFILCOUNTRY( Ecuador ) AND SUBJAREA(COMP) AND PUBYEAR < 2019

The previous query was executed in early 2020 and, 3411 documents were
retrieved, 6431 unique authors and 844 unique sources. From the list of authors,
we used the Author Retrieval API to get relevant data of authors such as their
metrics and alternate names. And, to get ranking information of sources, we
download data from the Scimago Journal Rank (SJR) site. Using the identifiers
of authors, documents and sources, the information finally was integrated.
To fill in the gender of the authors we made updates based on the most
popular names in the database. Then, to update missing data, we used the
Gender API which provides the gender probability according to the first name.
For authors with an incomplete name, like initials, or with a probability under
70% of confidence, the gender was set with NA value. Finally, we found that
71.7% of the authors are men, 24.6% are women, and for 3.7% of the authors,
their gender could not be determined. This result could be related with the
diﬀerence of percentage of female and male graduation of tertiary ICT programs.
According to data of the UNESCO Institute for Statistics, from 2014 to 2016,
the percentage of graduated were 2.4%, 3.3% y 2.9% for women and 6.5%, 7.4%
y 6.7% for men.

2.2 Scope definition and data analysis

From collected information, we analyze some variables and defined the scope of
this study based on the following criteria.

– Publication date. The first published document dates from 1986. From this
year until 2006, the annual production was minimal, i.e. around five papers
per year. From 2013, the production rise to more than 50 documents per
year. Considering the low production of the first years, we focus on the
analysis of documents indexed between 2009 and 2018. Applying the filter
for the selected period, 3363 documents were obtained, 48 documents minus
of those retrieved from the original query.
– Document type. Regarding the distribution of documents by type, three
categories had few publications, Book (2), Errata (1), and Note (1). These
types were discarded, therefore, we analyze 3359 documents.

Table 1 shows the number of documents per year and per type. From this
figure, two relevant facts are inferred 1) in the last couple of years (2017-2018),
2166 documents were indexed, i.e. 64% of Ecuadorian production in CS; and 2)
of each 5 documents indexed by Scopus, 4 are conferences, and 1 is an article.
This last data stats that CS researchers prefer conference proceedings as the
main source for publishing.

Table 1. Total of documents by year and type

Group of year Total Document type Total
2019 & 2010 46 Review 18
2011 & 2012 68 Editorial 21
2013 & 2014 163 Chapter 37
2015 & 2016 920 Article 579
2017 & 2018 2.166 Conference Paper 2.704

From the data collected, two dimensions led the analysis. For each dimension,
the following topics are studied:

1. Annual production and participation rates by gender:
– Production of leading authors
– Participation average rates
2. Publishing patterns by document type and impact:
– Production of leading authors by document type
– General distribution of leading authors
– Annual production by quartile
– Annual production by quartile and H index

To analyze the first topic of each dimension, we assume that leading authors
are those who have the first authorship. According to [11] the first author is
who led the research or who contributed the most to the research on which the
publication is based.

3 Results

In this section, we present the results for each analyzed topic. We use the Tableau
tool to create visualizations.

3.1 Annual production and participation rates by gender

Production of leading authors: Fig. 1 shows the annual evolution of the
scholarly production in CS by gender.

Fig. 1. Papers by year and first author gender

The stacked area chart shows that both women and men have increased
their production since 2014. But, the growth rate of publications led by male
(M) authors is higher than those achieved by female (F) authors. In the last
year analyzed (2018), of every 4 indexed documents, 3 have a man as the first
author, and only 1 paper has a woman as the first author.

Participation average rates: Table 2 presents the annual average of authors
per document, considering: 1) the total number of authors by gender, 2) the
total number of authors by gender when a woman is the first author, 3) the
total number of authors by gender when a man is the first author. As we can
see, when a woman is the first author, co-authors tends to distribute equitably
by gender. Whereas when a man is the fist author, the presence of women is
reduced.

Table 2. Annual average of authors per document

Annual average of authors
Year All papers Female leading author Male leading author
Total F M Total F M Total F M
2009 3,2 0,6 2,9 1,5 1,0 0,5 3,4 0,5 3,2
2010 3,2 0,6 3,0 3,4 1,8 1,8 3,2 0,4 3,3
2011 3,1 0,5 3,2 3,0 2,0 1,7 3,1 0,3 3,4
2012 3,5 0,9 2,8 3,5 1,8 2,1 3,6 0,6 3,1
2013 3,5 0,9 2,8 3,7 2,0 2,2 3,4 0,6 3,0
2014 3,9 1,1 3,2 3,8 1,9 2,2 4,0 0,6 3,6
2015 3,9 1,0 3,3 3,4 1,9 1,9 4,0 0,8 3,7
2016 4,2 1,0 3,5 3,9 2,0 2,2 4,3 0,7 3,9
2017 4,2 1,0 3,5 3,9 2,1 2,1 4,2 0,7 3,9
2018 4,2 1,1 3,4 4,0 2,0 2,4 4,2 0,8 3,8
Average 3,7 0,9 3,2 3,4 1,9 1,9 3,7 0,6 3,5

3.2 Publishing patterns by document type and impact

Production of leading authors by document type: Table 3 resumes the
type of document published considering the gender of leading authors. Article,
Conference paper and Editorial show similar percentages of participation for
both genders. On the other hand, in Chapter and Review, the participation of
women almost duplicate men. Furthermore, authors regardless of gender, have
preference to send theirs works to a conference, out of every 10 documents, 8
are in this category.

Table 3. Document type by gender of leading author

Gender of leading author
Document type Total (%)
F (%) M (%)
Article 17,2 16,3 17,5
Chapter 1,1 1,7 0,9
Conference Paper 80,5 80,6 80,5
Editorial 0,6 0,5 0,7
Review 0,5 0,9 0,4
General distribution of leading authors by quartile rank: Table 4 shows
the percentages of publication in articles and conference papers according the
gender and the quartile of publication. The ranking of journals are divided into
four quartiles: Q1, Q2, Q3 and Q4. Journals with the highest rank are in the
Q1 that represents the top 25% of journals. Authors, regardless the gender, have
similar preferences considering the quartile of the publications. 21% of authors
prefer to publish in articles in quartile Q1. Most of the woman (26%) prefer to
publish in conferences papers in quartile Q3 and 21% prefer quartile Q2. Most
of the men (27%) prefer to publish in conferences papers in quartile Q2 and 21%
prefer quartile Q3.

Table 4. Contributions by quartile

Document type
First author Best Quartile
Article Conference Paper Total
Q1 21,72% 0,71% 22,42%
Q2 12,02% 20,71% 32,73%
F
Q3 13,43% 26,36% 39,80%
Q4 4,24% 0,81% 5,05%
Q1 21,15% 0,89% 22,04%
Q2 13,17% 26,93% 40,09%
M
Q3 12,25% 21,26% 33,51%
Q4 3,57% 0,78% 4,35%

Annual production by quartile: This topic was analyzed considering 565
articles and 479 conference papers because not all of them are published in
sources with information from SJR.
Fig. 2 and Fig. 3 show articles and conferences paper published since 2009
to 2018 by quartiles, respectively.
Both women and men, have the same interest in publish articles in Q1, as
we can see in Fig. 2, there is a constant increase of publication in this quartile.
A possible diﬀerence between both genders is that women have published more
articles Q3 in 2018. Considering conference papers, both genders show a similar
pattern, the percentage of authors publishing in conference papers in Q3 have
increased considerably in the last year.

Annual production by quartile and h-index: Fig. 4 and Fig. 5 show articles
and conferences paper by quartiles and h-index, respectively. Considering the
articles, most of the them have low h-index, but in conference papers most of
the authors are looking for publishing in conferences of Q2 with high h-index.
Finally, after analyzing the six topics associated with the publication patterns
of leading authors, we have been able to confirm that men and women share
similar preferences when choosing the type of document, and the impact of
Fig. 2. Articles by year and quartile

Fig. 3. Conference papers by year and quartile
Fig. 4. Articles papers by quartiles and h-index

Fig. 5. Conference papers by quartile and h-index
the sources. However, there are two substantial diﬀerences 1) in the proportion
of men and women who publish as the first author, and 2) in the proportion
of co-authors by gender. Co-author groups in which a woman leads are more
homogeneous than those groups led by men.

4 Conclusion

In recent years, the scholarly publishing of Ecuadorian authors began to position
itself as a common practice among researchers. Computer Science production is
one of the promising in the country. Currently, the CS field ranks the third place
of the Ecuadorian scientific production indexed by Scopus.
In this study, we have found that there are diﬀerences in scientific production
by gender. Although the publication rate grows every year, those leading by men
are three times more than the leading by women. Of the 3359 documents indexed
by Scopus, 76% of the first authors are men, 23% are women and 1% represent
people with unidentified gender. This means that of four documents published,
three have a man as the first author, and only one is led by a woman. Despite
of this, women as first authors tend to work with equitable groups of co-authors
of both genders.
Furthermore, in order to determine the eﬀects of the low participation of
women in scientific production in the area, we evaluated some publishing pat-
terns considering two measures associated with the quality of the sources: quar-
tiles and h-index. With regard to the quality of conference papers, both genders
are inclined to quartiles Q2 and Q3. Also, since 2016, both genders have increased
article publications in Q1. It is also evident that most articles and conference
papers of both genders have a low h-index.
As a result, despite the great diﬀerence in number of female and male re-
searchers that leading scientific production, both genders have the same behav-
iors, in the sense that in recent years they have sought to increase their publica-
tions. Therefore, if women had greater participation in the scientific production
of the country, they could also improve their visibility and impact.
The findings found in this study encourage us to continue working on this
topic. In fact, this study is part of a larger initiative that tries to analyze gender
bias in Ecuadorian research. Currently, we are analyzing other indicators related
to the impact of researchers and papers. Gathering evidence on how women
publish in Computer Science can help government, and educational and research
institutions establish mentoring, support, and promotion programs to improve
women’s participation.

Acknowledgments

This work was partially supported by the scholarships provided by the Secretarı́a
Nacional de Educación Superior, Ciencia y Tecnologı́a of Ecuador (SENESCYT).
References
1. Aksnes, D.W., Rorstad, K., Piro, F., Sivertsen, G.: Are female researchers less
cited? A large-scale study of Norwegian scientists. Journal of the American Society
for Information Science and Technology 62(4), 628–636 (2011)
2. Bravo-Hermsdorﬀ, G., Felso, V., Ray, E., Gunderson, L.M., Helander, M.E.,
Maria, J., Niv, Y.: Gender and collaboration patterns in a temporal sci-
entific authorship network. Applied Network Science 4(1), 112 (2019).
https://doi.org/10.1007/s41109-019-0214-4
3. Bhrer, S., Frietsch, R.: How do public investments in gender equal-
ity initiatives and publication patterns interrelate? the case of
germany. Evaluation and Program Planning 79, 101752 (2020).
https://doi.org/https://doi.org/10.1016/j.evalprogplan.2019.101752
4. Dabas, B., Kumar, S.: Research output of indian women scientists in the field of
physics and astronomy: A scientometrics study. Library Philosophy and Practice
2018 (11 2018)
5. Garg, K., Kumar, S.: Scientometric profile of indian scientific output in life sciences
with a focus on the contributions of women scientists. Scientometrics 98(3) (03
2014). https://doi.org/10.1007/s11192-013-1107-4
6. Huang, J., Gates, A.J., Sinatra, R., Barabási, A.: Historical comparison of
gender inequality in scientific careers across countries and disciplines. CoRR
abs/1907.04103 (2019)
7. Isfandyari-Moghaddam, A., Hasanzadeh, M.: A study of factors inhibiting research
productivity of Iranian women in ISI. Scientometrics 95(2), 797–815 (May 2013).
https://doi.org/10.1007/s11192-013-0980-1
8. Larivire, V., Ni, C., Gingras, Y., Cronin, B., Sugimoto, C.R.: Bibliometrics: Global
gender disparities in science. Nature News 504(7479), 211 (2013)
9. Mihaljevi-Brandt, H., Santamara, L., Tullney, M.: The eﬀect of gender in the publi-
cation patterns in mathematics. PLoS One 11(10) (2016), publisher: Public Library
of Science
10. Nourmohammadi, H., Hodaei, F.: Perspective of iranian womens scientific produc-
tion in high priority fields of science and technology. Scientometrics 98 (02 2014).
https://doi.org/10.1007/s11192-013-1098-1
11. Orellana, D., Hermida, M.A., Osorio, P.: Lineamientos para Autorı́a de Publica-
ciones Cientı́ficas. Coyuntura 22(1), 81–90 (2018)
12. Puuska, H.M.: Eﬀects of scholars gender and professional position on publishing
productivity in diﬀerent publication types. Analysis of a Finnish university. Scien-
tometrics 82(2), 419–437 (2010)
13. Teele, D.L., Thelen, K.: Gender in the journals: Publication patterns in political
science. PS: Political Science & Politics 50(2), 433–447 (2017), publisher: Cam-
bridge University Press
14. UIS: Women in Science, http://uis.unesco.org/en/topic/women-science/
15. Vela, B., Cceres, P., Cavero, J.M.: Participation of women in soft-
ware engineering publications. Scientometrics 93(3), 661–679 (2012).
https://doi.org/10.1007/s11192-012-0774-x