2024 IEEE/ACM Workshop on Gender Equality, Diversity, and Inclusion in Software Engineering (GEICSE)

The Second Round: Diverse Paths Towards Software Engineering

Sonja M. Hyrynsalmi Ella Peltonen
[email protected] [email protected]
LUT University University of Oulu
LUT, Finland Oulu, Finland

Fanny Vainionpää Sami Hyrynsalmi

[email protected] [email protected]
University of Oulu LUT University
Oulu, Finland LUT, Finland
ABSTRACT
In the extant literature, there has been discussion on the drivers
and motivations of minorities to enter the software industry. For
example, universities have invested in more diverse imagery for
years to attract a more diverse pool of students. However, in our
research, we consider whether we understand why students choose
their current major and how they did in the beginning decided to
apply to study software engineering. We were also interested in
learning if there could be some signs that would help us in mar-
keting to get more women into tech. We approached the topic via
an online survey (N = 78) sent to the university students of soft- Figure 1: Some example pictures used to market software
ware engineering in Finland. Our results show that, on average, engineering (or ICT field as a whole) for BSc degree studies.
women apply later to software engineering studies than men, with Pictures are quoted from the Universities’ online materials
statistically significant differences between genders. Additionally, and StudyInfo, the governmental application website.
we found that marketing actions have different impacts based on
gender: personal guidance in live events or platforms is most influ-
ential for women, whereas teachers and social media have a more
significant impact on men. The results also indicate two main paths
into the field: the traditional linear educational pathway and the 1 INTRODUCTION
adult career change pathway, each significantly varying by gender. Research has shown that the primary factors influencing students’
choice of study program in information systems are career-oriented.
CCS CONCEPTS The most significant factor for students when selecting a study
• Social and professional topics → Information science educa- program was identified as job availability, closely followed by job
tion; Computer and information systems training; Software security, career opportunities, and the prospect of engaging in
engineering education; Gender; • Human-centered comput- work assignments. [1] It has also been highlighted that students
ing → Empirical studies in ubiquitous and mobile computing. in Information and Communication Technology (ICT) generally
share similar perceptions of the field with students from other areas,
KEYWORDS with the notable exception that they view ICT as more creative
Software Engineering Education, Gender, University Student Ad- and people-oriented than students in the other fields. This insight
mission underscores the importance of highlighting the creative aspects
of software engineering to potentially attract a broader and more
ACM Reference Format: diverse student body, including increasing the number of women
Sonja M. Hyrynsalmi, Ella Peltonen, Fanny Vainionpää, and Sami Hyryn- in ICT studies. [33].
salmi. 2024. The Second Round: Diverse Paths Towards Software Engineer-
However, although more attention has been paid to the more
ing. In 2024 ACM/IEEE Workshop on Gender Equality, Diversity, and Inclusion
in Software Engineering (GE@ICSE ’24), April 20, 2024, Lisbon, Portugal. ACM,
diverse marketing, for example, the images used in the university
New York, NY, USA, 8 pages. https://doi.org/10.1145/3643785.3648494 websites and marketing (Figure 1), the number of women in soft-
ware engineering studies has increased slowly. This also affects
the representation of women in software engineering, as fewer
women are transitioning to working life [22]. Women face different
This work licensed under Creative Commons Attribution International 4.0 License. socio-cultural challenges (such as work-life balance issues, impos-
tor syndrome and a lack of recognition and peer equality) in their
GE@ICSE ’24, April 20, 2024, Lisbon, Portugal studies and working life [30, 46] and attracting women to the soft-
© 2024 Copyright held by the owner/author(s).
ACM ISBN 979-8-4007-0575-5/24/04 ware engineering field is that way harder, although women would
https://doi.org/10.1145/3643785.3648494 have the same skills as men.

29
GE@ICSE ’24, April 20, 2024, Lisbon, Portugal Hyrynsalmi et al.

To get a more timely understanding of the gender differences and the knowledge-sharing culture as a whole [48] and boost in-
between different marketing strategies in software engineering ed- novation and creativity activities [27]. There is a common belief
ucation, we conducted an online survey (n = 78) for university-level that women would pursue IT because of passion and men because
software engineering students in Finland. In this research paper, we of material concerns[42]. However, some studies also show that
provide preliminary insights into our data, which will be extended female and male professionals value actually similar features in
with further interviews and more detailed questionnaires about their software engineering jobs: good work-family balance, salary
reasons behind applying to software engineering programs and and other benefits, and career advancement opportunities [24, 40].
marketing efforts that may have impacted the applicants’ decisions. Still, as women are still underrepresented in the field, even the
With a preliminary answer pool consisting of half-female and half- best diversity recruitment practices [15] do not help if there is
male attendants, we can present the first comparison efforts in this no population from where to recruit. Thus, universities and other
paper. higher education providers are also challenged to attract students
This study addresses the following research questions (RQ): of diverse backgrounds and genders to their programs.
Some evidence shows that female students who do not consider
RQ1 What are the perceived reasons for selecting software en- a software career for themselves still value IT highly [29]. Further-
gineering university studies, and how do they differ between more, the female students attending the ICT programmes seem
genders? to perceive it as a prospective profession, even if they may lack
RQ2 When, where, and how do people of different genders find self-esteem compared to their male counterparts [34]. Diminish-
information and influence the decision to apply for software ing gendered stereotypes in the ICT careers and underlining the
engineering degrees? required "soft" skills seem to be the first steps towards higher par-
The remaining sections are structured as follows: Section 2 re- ticipation of women in the tech industry. Still, the evidence also
views related work and central concepts. Section 3 presents the suggests that these actions alone may not be enough [44]. Even if
empirical research approach, and the results are shown in Section 4. the prospective female students see the opportunities in the ICT
Discussion and conclusions are given in Sections 5 and 6. field, it may require more to get them actually apply to the ICT pro-
grammes. At the same time, the faculty and university teachers tend
to see their programmes enough aware of diversity practices [25].
2 BACKGROUND Still, the number of female students is increasing but too slowly.
2.1 Women studying STEM
Why don’t more females choose the technology field? The ques- 2.3 Where to find the female software engineers
tion is discussed in research publications and in many university
marketing meetings. The reasons have been found in stereotypical The problem described here is long known, and successful projects
thinking of female and male career paths, which can already man- have generated insights and ideas for more diverse curricula [45].
ifest in a lack of female teachers in the mathematics and natural There have been hackathons [32], actions towards women career-
sciences [5, 9]. Another possible reason is seen as a non-technical changers [7, 24] and other, often interdisciplinary, initiatives [19, 21]
self-assessment of female students, guided by expectations of soci- involving more female students in STEM and software engineering
ety and families [17, 18, 43]. However, when they complete their studies. These activities have a significant place in the field of
studies, female graduates have little to no difference in skills from raising awareness of technology professions. However, university
their male counterparts [47], underlining how important it is to get student recruitment is an ongoing, long-term process different
that skilled workforce interested in the technology industry. from specialising projects. It is often designed and implemented
Successful marketing to gather more female students in the by specialised marketing professionals instead of STEM faculty
STEM (Science, Technology, Engineering, and Mathematics) field is whose work time is effectively reserved for research and education.
often seen as a representation problem. Solutions include increasing Resourced used for marketing actions, especially in publicly funded
the number of females in images used in student marketing cam- universities (like in Finland, where this study takes place), are often
paigns and publishing stories featuring female technology profes- limited and shared with all the other disciplines. These resource
sionals as role models [23, 31]. Other suggestions include organising limitations often make hands-on activities and personal connections
more hands-on events, such as all-female coding workshops and in the field very challenging.
tech seminars for non-technology-oriented students. Indeed, such University applicant attraction efforts can include multiple differ-
hands-on activities seem to arouse interest towards the STEM ca- ent marketing strategies [8] and branding activities [4, 35]. Physical
reer [16, 17, 24]. Equally, access to a "real" female STEM professional recruitment, i.e. in-campus events and study fairs, is often reported
increases female students’ self-esteem regarding their potential and as efficient applicant marketing together with the university web-
self-perceived skills towards the STEM career [20, 26]. sites and brochures [14] and student-written web content [38] tar-
geted for the applicants. In addition, universities and other higher
education providers are attracted by branding a social media pres-
2.2 Software Engineering in diversity crisis ence [14] for marketing and recruitment efforts. These can include
Diversity in software engineering has explicitly called to be in a hosting university-specific pages or feeds [3] but also ads in video
crisis [2]. Gender-diverse teams have been reported to perform streaming services such as Youtube [13] and even utilising services
better in software engineering projects [12, 37]. Thus, getting more of social media influencers [36, 39]. Social media campaigns often
females present in the field could improve the quality of the results target younger generations through social media channels such

30
The Second Round: Diverse Paths Towards Software Engineering GE@ICSE ’24, April 20, 2024, Lisbon, Portugal

as TikTok, Snapchat, and Instagram [41]. More "adult" channels The survey was designed and implemented in Finnish as this
such as LinkedIn may be dismissed [41] even if, as we later show in research focused on the effects of marketing activities in Finland.
this paper, several students apply in their first or second university While most of the studied universities offer either Bachelor’s or
programme only later in life. The same risk lies with the student Master’s programmes taught in English, the marketing strategy
recruitment events as they are usually targeted at the local high and activities for those programmes often differ from the marketing
school students. to Finnish-speaking citizens, often applying for their first study
position in a higher-education institution.
3 RESEARCH APPROACH The designed survey was tested with two students and reviewed
by one software engineering programme leader. Based on their
3.1 Research process
feedback, some questions were clarified and reworded to make
This study is carried out as a survey of university students majoring them unambiguous.
in software engineering at Finnish universities as this exploratory Soon after the study was published, it was noted that the univer-
study aims to generate new insights into the topic. The chosen sity’s website was a frequently appearing option- among the free
research strategy follows a mixed-method approach. The research answers for the questions where the respondents sought further
process used in this study consists of two major phases: preparation information regarding the study programmes. This option was then
and execution. The first major phase consists of literature and added among the alternatives in the respective question.
background study, identifying the relevant institutions in Finland,
and contacting their education programme managers.
First, after the literature study, we listed all universities in Fin-
3.3 Participant recruitment
land that have a major in software engineering. In this phase of The survey is designed for university students majoring in soft-
study, polytechnics were omitted. Out of the 13 universities in ware engineering at Finnish universities. Especially the focus is
Finland, nine universities are offering degree studies in software on understanding the impact of marketing activities focused on
engineering (Aalto University, LUT University, Tampere University, Finnish-speaking candidates and their influences on different gen-
University of Eastern Finland, University of Helsinki, University ders. The study is designed as an open public subscription survey;
of Jyväskylä, University of Oulu, University of Turku, and Åbo that is, the objective is to gather as many respondents belonging to
Akademi University). We contacted software engineering profes- the target population as possible.
sors in each university and inquired about information regarding The size of the target population is hard to estimate as the official
the persons responsible for marketing or leading software engi- role of the software engineering education programmes varies a
neering programs in their institution. lot between different institutions. However, according to statistics
During August-September 2023, we held separate meetings with from The Ministry of Education and Culture and the Finnish Na-
each university’s responsible marketing programme managers. Dur- tional Agency for Education, there are 7,734 students in Finnish
ing the discussion, we confirmed their institutions’ software engi- universities in different programmes in the field of Information
neering education programmes and invited them to forward our Technology. This number, however, includes also, for instance, com-
questionnaire to their students. All universities agreed to take part puter engineering, telecommunication, and data science majors. As
in the study. a rough estimate, around 1,000–2,000 students major in software
The second major phase consisted of designing, conducting and engineering.
analysing the survey of the software engineering major students. The participant recruitment was done via two main channels:
Each university’s corresponding person sent the survey directly First and foremost, each university’s software engineering program
to the software engineering majors in their university in August- leader, or corresponding, was asked to send the invitation to the
September 2023. Some universities also sent a remainder later, usu- students of their programmes. All universities agreed to do this.
ally in mid-October. Finally, the analysis and reporting were carried Second, the survey questionnaire was distributed through social
out during November-December 2023. media channels such as through various student organisations’ X
(Twitter), LinkedIn, and Instagram accounts.
3.2 Survey design
The survey instrument was designed in August 2023. The survey 3.4 Analysis
was designed to be exploratory in nature as it aims to generate in- The survey consisted of both qualitative and quantitative questions,
sights into the topic at hand. The questions relevant to this study’s and therefore, a mixed-method approach for the analysis was se-
goals were characterised and improved in meetings with the au- lected. Mixed methods are suitable for this study as the research
thors. The used questions include free answer questions, ranking objectives, on the one hand, involve both qualitative (e.g., RQ1) and
questions, and question asking to select the most suitable option. quantitative (e.g., RQ2) aspects; in addition, the research approach
The survey was implemented using the Webropol online question- in this study is exploratory, thus benefiting from a mixed-method
naire survey tool. approach.
The selected 22 questions were grouped into three main themat- The quantitative answers and the corresponding open answers
ical groups: (i) experiences and reasons to select software engineer- were considered to provide qualitative insights into the numbers.
ing, (ii) experiences on marketing activities, and (iii) background Quantitative answers were divided into female and male groups,
information. The survey was designed to be anonymous; however, and the statistical independence between these groups was tested
the basic background information (age, gender, etc.) was asked. with suitable statistical tests. The tests used were the Wilcoxon

31
GE@ICSE ’24, April 20, 2024, Lisbon, Portugal Hyrynsalmi et al.

Women Male
18-20 5.3% 5.7%
21-30 23.7% 54.3%
31-40 52.6% 28.6%
41-50 15.8% 8.6%
51-60 2.6% 2.8%
Table 1: Age distribution of the questionnaire participants by
gender. The major age groups are highlighted. All answers
are between 18 and 60 years old. Figure 2: Answer distribution to the question "Do you remem-
ber when you decided to apply to study computer science?"
(p = 4.3248e-23)

rank sum test, which is a nonparametric test for two populations

when samples are independent, and the Kruskal-Wallis test, which 4.2 When do they apply for the software
is a nonparametric version of classical one-way ANOVA and an
extension of the Wilcoxon rank sum test to more than two groups.
engineering studies?
For the qualitative analysis, we used thematic analysis as guided "Do you remember when you decided to apply to study soft-
by Braun and Clarke [6]. We included the 13 open-ended questions ware engineering?" This question aims to survey when students
used in the survey for the qualitative analysis. The analysis was first thought of studying the ICT field, or software engineering
done in the six steps: (i) familiarising with the data, (ii) initial codes, more specifically, compared to their respective school history. The
(iii) theme identification, (iv) reviewing themes, (v) naming themes, options were given as primary school (age 7 to 12 in the Finnish
and (vi) reporting. Finally, the qualitative and quantitative analysis school system), secondary school (age 13 to 15), high school (or
results are combined into a simplified model. corresponding vocational school, typically age 15 to 18), university
of applied sciences (after age 18; often an option for the university
4 RESULTS studies in the first place), or later in the adulthood, after finish-
ing the first formal education. The option "other" was given but
4.1 Demographic distribution received only an answer "during the gap year".
The demographic information was collected in the following cate- The results of this question are shown in Figure 2. The answers
gories. Gender was requested as a categorical variable female, male, are divided by gender. Our analysis with the Wilcoxon rank sum test
other, and do not wish to disclose. Similarly, the age group was a shows that there is a statistical difference between female and male
categorical variable in decades, i.e. 18 – 20 years old as a fresh high groups (𝑝 = 4.3248𝑒 − 23); female participants decided to achieve
school graduate, 21 – 30 years old as a little older student, and so software engineering degrees later in their lives, whereas male
on. In addition, previous work experience, previously completed participants already figured their future field in the high school
degrees, study degree level (both BSc and MSc or MSc programme age. Our study attracted many participants whose decision to study
options) and the name of the home university were requested. software engineering was made later in life, which is also visible in
The questionnaire received 78 answers, from which 38 partici- the age distribution (Table 1). Especially many female participants
pants identified as a woman (50.7%), 35 as a male (46.7%), one other, belong to the age group 31–40 years old; at the same time, 68.4%
and one did not prefer to disclose. Due to only two participants of our female participants decided to achieve a degree in the field
reporting something other than female or male, this study focuses only after their first formal education.
on these two gender categories only and later discusses only par- In open responses, students highlighted different reasons for
ticipants identifying as female or male (N=76). The age distribution applying to software engineering education and their motivation for
is given in Table 1, showing that, on average, the male participants it. For example, some female respondents reported a longstanding
were younger than the female participants. Around half of the par- interest in technology. Still, for various reasons, they were guided
ticipants had no previous work experience before beginning their towards different fields in their studies, and only later did they
university degree. 32.4% of women had previous ICT-related work decide to pursue software engineering. This deviation could have
experience and 21.6% some other type of work experience. Similarly, been due to a lack of study counselling or perhaps the influence of
37.2% of males had ICT-related work experience, and 11.4% had a family member:
some other field experience. "For me, the main reason to study computer science was a very long-
63.9% of female participants and 68.6% of male participants had standing desire to pursue a technical degree. I considered applying
a study right for both BSc and MSc degrees. 36.1% of female par- for IT studies in high school, but my guidance counsellor and family
ticipants and 31.4% of male participants only studied in an MSc encouraged me apply for business studies." Woman, 31-40 years old
programme, usually indicating that this is their second degree (ei- Some respondents illustrated a journey of rediscovery and over-
ther completed BSc in another programme or in a university of coming initial apprehensions about software engineering. Although
applied sciences. The questionnaire was sent to students of nine some students had previously considered software engineering to be
national universities where software engineering (or a similar field an interesting field, they initially believed the studies to be too chal-
with a different name) is taught. The respondents came from seven lenging and assumed coding skills as a prerequisite. However, their
different universities. exposure to IT work in a different professional setting reignited

32
The Second Round: Diverse Paths Towards Software Engineering GE@ICSE ’24, April 20, 2024, Lisbon, Portugal

Figure 3: Answer distribution to the question "What influ- Figure 4: Answer distribution to the question "Where did you
enced you to apply to your university?" (p = 0.0073) find out about the programme you got into?" (p = 0.0137)

not limited to technical fields. Additionally, the possibility for re-

their interest and led them to reconsider software engineering as a mote studies and the flexibility of a schedule-independent learning
viable and desirable educational path: format were cited as significant advantages:
"My interest was sparked while working in a different field when "The primary reasons for choosing are: 1. Suitable threshold criteria
I had the opportunity to see up close the work my colleagues in IT for master’s selection -> can apply with a lower academic background,
were doing. Software Engineering had already crossed my mind, but not from a technical field. 2. Studies are handled entirely remotely. 3.
I thought the studies would be too difficult then and that practical Studies are almost entirely independent of schedule." Woman, 41-50
coding skills were required beforehand." Man, 21-30 years old years old

4.3 What influences on the decision to apply? 4.4 What was the influence of university
Where did you find out about the programme you were ac- marketing on application decisions?
cepted into? What influenced you to apply to your univer- Where did you find out about the programme you got into?
sity? This question aims to understand the influence of different This question aims to find out resources where the current students
pre-defined features on the decision-making of where to apply gathered information about their upcoming studies. The pre-defined
(programme or university). The question was answered by ranking features on the information sources were given as different types of
suitable options by importance. Not all the options had to be picked, visits (either into the university or different types of student fairs, or
and an open answer was provided (discussed later in this paper). a visitor from the university coming into the school environment),
Figure 3 summarises the results as average grades for each option, social media (either an ad or more personal post), and other more
divided into gender-specific categories. The statistical analysis with close-tied communication forms such as hearing/discussing with a
the Kruskal-Wallis test returns a statistical difference between the teacher, relative, or friend. Also, the University’s presence in general
answers of female and male participant groups (𝑝 = 0.0073). How- news and the University’s websites were included. The question
ever, the differences are subtle: both gender categories preferred was answered by ranking suitable options in order of importance.
similar topics, such as the cheap cost of living in the university area, Not all the options had to be picked, and an open answer was
the number of job opportunities in the region, and the university’s provided (discussed later in this paper).
size in their answers, only with different weights. Figure 4 summarises the answers to this question, divided by gen-
In the open responses, students revealed more information about der. The Kruskal-Wallis test returns a statistical difference between
their decision-making process. The role of social media was pointed the answers of female and male participant groups (𝑝 = 0.0137).
out in many answers. Respondents noted seeing posts on platforms Beyond effective social media marketing, the presence of easily
like Instagram and Facebook, and LinkedIn profiles of current stu- accessible and experienced faculty members or role models was
dents, which highlighted the professional opportunities available identified as crucial and influential. Typically, a single advertise-
post-graduation: ment alone was not a decisive factor; instead, it was the combina-
"I saw posts on Instagram and Facebook. Additionally, I saw people tion of multi-channel and diverse voices in communication that
on LinkedIn.com who study in the same program and the kinds of contributed to making an informed decision:
jobs they have." Woman, 31-40 year old "I first came across the advertisement for the program I’m in now
Key factors influencing the decision to enrol on some programs in a Facebook group, where a faculty member of my university en-
included the program’s inclusive eligibility criteria, allowing ap- couraged people to apply (the group was, if I recall correctly, a coder
plications from individuals with diverse academic backgrounds, community group). This led me to explore the university’s website

33
GE@ICSE ’24, April 20, 2024, Lisbon, Portugal Hyrynsalmi et al.

Figure 5: Diverse Paths to IT: Linear Educational Pathway and Adult Career Change Path identified in thematic analysis

and the specific program’s page, and eventually, I contacted the fac- advanced mathematics [syllabus] would provide a good foundation
ulty member directly before applying to the program." Woman, 31-40 for succeeding in further studies." Woman, 51-60 years old
years old. In Linear Educational Pathway, the main drivers were positive
The influence of close relationships in decision-making was par- experiences around technology. However, the thematic analysis
ticularly evident among participants, both in close-ended questions revealed that experiences were also crucial for those who made
and voluntary open-ended questions, when asked about finding career changes to software engineering in adult life — the second
information about the programme: round in this research. In analysis, experiences of failure in tech-
"I generally don’t use social media and don’t visit forums either. nology subjects or beliefs such as that the software engineering
I try to block advertising by all means. A friend of mine previously field requires deep mathematics occurred in the answers of adult
studied at the same university, and I was interested in seeing what it’s career change path respondents. These respondents saw their past
really like to study there. However, I gathered most of my information working experience as valuable for the IT sector, desired a better
from the university’s website." Man, 31-40 years old work-life (such as work with remote work possibilities or flexi-
bility), or were looking for possibilities for personal growth and
the possibility to update their skills. What was common was that
4.5 The Impact of Education and Work they got positive experience with technology topics in their career,
Experience to the Path Towards IT usually via their work duties around IT. Those positive experiences
In the thematic analysis of the answers to the questions "How has were strong enough to overcome previous bad experiences during
your previous work experience influenced your chosen field high school and supported the career change to IT.
and specialisation?" and "How has your educational journey "I see that my previous study path had very little effect on my
shaped your career decisions so far?", we revealed a multifac- current choice because I come from a high school where studies in
eted narrative. The results form a life-course [11] inspired theory the field of technology were not really advertised at all for female
(Figure 5) of the "Linear Educational Pathway" and "Adult Career people, and I ended up applying and getting a study place at a business
Change Path" forming diverse paths towards software engineering school after high school. I think my choices have been influenced more
education. by the strong role models I found in adulthood as well as my own
Linear Educational Pathway presents a more ’traditional’ path, non-technical career in the IT industry, which eventually made me
where individuals move directly from the high school to the uni- realise that the IT industry and the studies in the field are really for
versity. In thematic analysis, these respondents were divided into everyone" Woman, 31-40 years old
two groups: Those who had a passion for technology and were What was similar between these two paths was that they both
studying maybe in some high school with a technical focus. Then shared an interest in technology, starting from a young age. In their
there are those who are more uncertain about where their passion decision to apply towards software engineering education, close
lies but decide to apply for the study position at the university, as acquaintances such as family members, teachers, or life partners
was expected of them. played an important role. In these paths, respondents also reported
"High school studies mainly influenced me in such a way that choos- that they were interested in pursuing software engineering, as
ing this field did not feel too scary. I presumed that [high school’s]

34
The Second Round: Diverse Paths Towards Software Engineering GE@ICSE ’24, April 20, 2024, Lisbon, Portugal

there has been an ongoing discussion in Finland about the labour Our study also reveals that personal guidance is crucial in at-
shortage problem in software engineering: tracting women to software engineering, while teachers and social
"(Study Path) Had no actual effect. The expert shortage in IT jobs media influence more men. This difference suggests a need for
has affected my decision much more." Man, 41-50 years old tailored outreach strategies. For women who may no longer have
This theory of diverse paths to IT sheds light on how the in- access to traditional educational guidance, such as high school
terplay of professional experiences and educational background counselling, in their adult career change path, alternative methods
uniquely positions individuals in their career paths. It is essential of personal outreach, like active engagement through various social
to recognise that while our study identifies two primary pathways media channels and events, could be effective. As a result that men
into IT studies – directly from high school and adult career change favour social media, there can be a marketing bias: There have been
– these paths do not reflect the unique experiences of every indi- previous findings that IT jobs were primarily targeted at men. In
vidual. Both paths, despite their differences, reveal crucial points that way, more men clicked the advertisements and applied for the
where paths divide from each other and in which points they share positions. [28] This can also be an issue in our case. Furthermore, in
similar influence points. Related to the previous findings from the our research, we found out that the presence of faculty members in
quantitative data, it may also be that these experiences of adult social media spaces, both for program promotion and for providing
career change paths are more frequently the experiences of women. accessible information about studies, is an effective channel and
way to encourage women to apply. This indicates that it matters
5 DISCUSSION how social media is used to attract more women to software engi-
neering — but it also raises a question about resources, as this kind
5.1 Key findings of marketing would require someone from the faculty.
We summarise our study’s key findings in the following: Our study presents findings that add a new dimension to the
ongoing discussion of attracting more women to software engi-
First, our study shows that, on average, women are applying
neering education. Our ’Diverse Paths to IT’ theory emphasises
later into software engineering studies, and the difference
that universities’ programme marketing strategies in the field of
between male and female answers is statistically significant.
software engineering should identify and support both paths. Espe-
Second, the most important rationales behind the university
cially for women, the "adult career change path" is significant, as
selection for software engineering studies are the size of
women in our research seek an IT career later in life, utilising their
the university, cheap cost of living, and the number of job
previous experiences and skills.
opportunities. In addition, the diverse job opportunities and
the flexibility of the studies are seen as important.
Third, the results show that the most affecting marketing ac- 5.2 Limitations and future studies
tions to women are personal guidance; however, teachers Naturally, there are limitations worth discussing. Age distribution
and social media are the most important for men. is skewed towards older students. As usual, those with strong opin-
Fourth, the results show two main, yet remarkably differen- ions or the topic is close to their heart are more willing to answer
tiating, paths to the field: the linear educational and adult the survey. This is also visible in our study, as we got half of the re-
career change paths. This concept helps to understand that sults from women, even if they represent only circa 25% of Finnish
there are several career paths in the field of software engi- university students. We will consider this in future studies and
neering, and they can be very different depending on gender repeat the questionnaire with a larger population. This moment
and life cycle stages. study focus only to the Finnish students, but can be repeated in the
global perspective later. The questionnaire was only shared with
Our study supports the findings that women are interested in
the student population that was indeed accepted into the university
software engineering. Still, they may sometimes have to do a sec-
studies; it would be interesting to study those who applied to the
ond round — first, go to one career and then make a career change
software engineering programs but did not get accepted.
to software engineering. The findings of our thematic analysis also
support the previous findings [22] that social and environmental
factors contribute to the underrepresentation of women in engi- 6 CONCLUSION
neering — not their lack of interest in mathematical skills. The This study addressed two research questions: why did Finnish soft-
stereotype image that the software engineering field requires deep ware engineering students select their current major, and how did
mathematical skills is still strong. they decide to apply to study software engineering? In addition,
The emphasis on the desire for a better work-life, flexibility, we studied whether there are differences between genders in these
and remote learning options when making the decision points to two questions. The survey results (n=78) showed that women often
a potential shift in priorities. This can be possibly influenced by decided later in their life to apply to study in software engineering
post-COVID-19 trends around remote and hybrid working, which than men. In addition, there are differences in the main influencing
have been favourable to women and to increasing the proportion factor driving the software engineering field; however, this might
of women and reducing the workload in the home, and sometimes be partially explained by the age differences between the male and
quite opposite [10]. This raises the question of whether there are female respondents. We summarised the key findings into a model
barriers on the path to technical fields, especially for women, that depicting two different paths to software engineering studies: the
could be mitigated through more flexible and accessible educational linear path and the career change path. The latter of these explains
models. This could be one future study angle. the name of this study: for a remarkable number of students in

35
GE@ICSE ’24, April 20, 2024, Lisbon, Portugal Hyrynsalmi et al.

Finnish universities, software engineering studies are the second finland. In 2019 IEEE/ACM 41st International Conference on Software Engineering:
round after completing the first degree in some other field. Companion Proceedings (ICSE-Companion). IEEE, 304–305.
[24] Sonja M Hyrynsalmi. 2019. The underrepresentation of women in the soft-
ware industry: thoughts from career-changing women. In 2019 IEEE/ACM 2nd
International Workshop on Gender Equality in Software Engineering (GE). IEEE.
ACKNOWLEDGMENTS [25] Sonja M. Hyrynsalmi. 2023. How Diversity and Inclusion Are Approached in
This research was supported by PHP Säätiö (PHP Holding), Grant Software Engineering University-level Teaching. In IEEE/ACM 4th Workshop on
Gender Equity, Diversity, and Inclusion in Software Engineering (GEICSE). 17–24.
20230012 [26] Regina Kelly, Oliver McGarr, Louise Lehane, and Sibel Erduran. 2019. STEM and
gender at university: focusing on Irish undergraduate female students’ percep-
tions. J. of Applied Research in Higher Education 11, 4 (2019), 770–787.
REFERENCES [27] Karina Kohl and Rafael Prikladnicki. 2022. Benefits and Difficulties of Gender Di-
[1] Ali Acilar and Øystein Sæbø. 2022. Exploring the differences between students in versity on Software Development Teams: A Qualitative Study. In XXXVI Brazilian
IS and other disciplines in the perceptions of factors influencing study program Symposium on Software Engineering. 21–30.
choice: A survey study in Norway. (2022). [28] Anja Lambrecht and Catherine Tucker. 2019. Algorithmic bias? An empirical
[2] Khaled Albusays, Pernille Bjorn, Laura Dabbish, Denae Ford, Emerson Murphy- study of apparent gender-based discrimination in the display of STEM career
Hill, Alexander Serebrenik, and Margaret-Anne Storey. 2021. The diversity crisis ads. Management science 65, 7 (2019), 2966–2981.
in software development. IEEE Software 38, 2 (2021), 19–25. [29] Katja Leiviskä and Mikko Siponen. 2010. Attitudes of sixth form female students
[3] Costas Assimakopoulos, Ioannis Antoniadis, Oliver G Kayas, and Dragana Dvizac. toward the IT field. ACM SIGCAS Computers and Society 40, 1 (2010), 34–49.
2017. Effective social media marketing strategy: Facebook as an opportunity for [30] Jane Margolis and Allan Fisher. 2002. Unlocking the clubhouse: Women in comput-
universities. Int. Journal of Retail & Distribution Management 45, 5 (2017). ing. MIT press.
[4] JH Beneke. 2011. Marketing the institution to prospective students-A review of [31] Donna Milgram. 2011. How to recruit women and girls to the science, technology,
brand (reputation) management in higher education. Int. journal of Business and engineering, and math (STEM) classroom. Technology and engineering teacher
Management 6, 1 (2011), 29. 71, 3 (2011).
[5] Martha Cecilia Bottia, Elizabeth Stearns, Roslyn Arlin Mickelson, Stephanie [32] Lavínia Paganini, Kiev Gama, Alexander Nolte, and Alexander Serebrenik.
Moller, and Lauren Valentino. 2015. Growing the roots of STEM majors: Female 2023. Opportunities and constraints of women-focused online hackathons. In
math and science high school faculty and the participation of students in STEM. IEEE/ACM 4th Workshop on Gender Equity, Diversity, and Inclusion in Software
Economics of Education Review 45 (2015), 14–27. Engineering (GEICSE). IEEE, 33–40.
[6] Virginia Braun and Victoria Clarke. 2006. Using thematic analysis in psychology. [33] J.M. Rajala, N. Iivari, M. Kinnula, D. Rajanen, and T. Molin-Juustila. 2022. Perceiv-
Qualitative Research in Psychology 3, 2 (2006), 77–101. ing ICT: Factors Influencing the Selection of Information Systems as a Major.
[7] Barbora Buhnova and Dita Prikrylova. 2019. Women want to learn tech: Lessons [34] Nevena Ranković, Elinda Kajo Mece, Mirjana Ivanović, Asya Stoyanova-
from the Czechitas education project. In 2019 IEEE/ACM 2nd International Work- Doycheva, Miloš Savić, and Dragica Ranković. 2019. Female Students’ Attitude
shop on Gender Equality in Software Engineering (GE). IEEE, 25–28. Towards Studying Informatics and Expectations for Future Career-Balkan Case.
[8] Mark Camilleri. 2020. Higher education marketing communications in the digital In 9th Balkan Conference on Informatics. 1–7.
era. In Strategic marketing of Higher education in Africa. Routledge, 77–95. [35] Philipp A Rauschnabel, Nina Krey, Barry J Babin, and Bjoern S Ivens. 2016. Brand
[9] Serena Canaan and Pierre Mouganie. 2023. The impact of advisor gender on management in higher education: the university brand personality scale. Journal
female students’ STEM enrollment and persistence. Journal of Human Resources of Business Research 69, 8 (2016), 3077–3086.
58, 2 (2023), 593–632. [36] Megan Rolfs and Josie Ahlquist. 2021. Opportunities for student influencer
[10] Linda L Carli. 2020. Women, gender equality and COVID-19. Gender in Manage- marketing strategies. Journal of Education Advancement & Marketing 5, 4 (2021).
ment: An Int.Journal 35, 7/8 (2020), 647–655. [37] Daniel Russo and Klaas-Jan Stol. 2020. Gender differences in personality traits of
[11] C. Castaño and J. Webster. 2011. Understanding women’s presence in ICT: The software engineers. IEEE Trans. on Software Engineering 48, 3 (2020), 819–834.
life course perspective. Int. J. Gender, Science & Technology 3, 2 (2011), 364–386. [38] J.K. Sandlin and E. Vallejo Peña. 2014. Building authenticity in social media
[12] Gemma Catolino, Fabio Palomba, Damian A Tamburri, Alexander Serebrenik, tools to recruit postsecondary students. Innovative Higher Education 39 (2014),
and Filomena Ferrucci. 2019. Gender diversity and women in software teams: 333–346.
How do they affect community smells?. In IEEE/ACM 41st Int. Conf. on Software [39] Peter Schneckenleitner and Johanna Thaler. 2020. The Use of Educational Influ-
Engineering: Software Engineering in Society (ICSE-SEIS). IEEE, 11–20. encers for Communication Activities in the Tertiary Education Sector in Austria.
[13] E. Constantinides and M.C Zinck Stagno. 2012. Higher education marketing: A Proceedings of INTCESS (2020), 524–531.
study on the impact of social media on study selection and university choice. Int. [40] Marvin Schuth, Prisca Brosi, and Isabell Welpe. 2018. Recruiting Women in IT: A
journal of technology and educational marketing (IJTEM) 2, 1 (2012), 41–58. Conjoint-Analysis Approach. In 51st Hawaii Int. Conf. on System Sciences.
[14] Efthymios Constantinides and Marc C Zinck Stagno. 2011. Potential of the social [41] Alison B Shields and Adam Peruta. 2019. Social media and the university decision.
media as instruments of higher education marketing: A segmentation study. Do prospective students really care? Journal of marketing for higher education
Journal of marketing for higher education 21, 1 (2011), 7–24. 29, 1 (2019), 67–83.
[15] Ella Dagan, Anita Sarma, Alison Chang, Sarah D’Angelo, Jill Dicker, and Emerson [42] Nur Soylu Yalcinkaya and Glenn Adams. 2020. A cultural psychological model of
Murphy-Hill. 2023. Building and Sustaining Ethnically, Racially, and Gender cross-national variation in gender gaps in STEM participation. Personality and
Diverse Software Engineering Teams: A Study at Google. In 31st ACM Joint Social Psychology Review 24, 4 (2020), 345–370.
European Software Engineering Conference and Symposium on the Foundations of [43] Noshin Tahsin, Nazmus Sakib Ahmed, Moumita Asad, and Kazi Sakib. 2022. Can
Software Engineering (ESEC/FSE 2023). Association for Computing Machinery, female underrepresentation in information technology be solved through an
New York, NY, USA, 631–643. awareness-based approach?. In 3rd Workshop on Gender Equality, Diversity, and
[16] Ismail Donmez. 2021. Impact of Out-of-School STEM Activities on STEM Career Inclusion in Software Engineering. 1–5.
Choices of Female Students. Eurasian Journal of educational research 91 (2021). [44] Hau-lin Tam, Angus Yuk-fung Chan, and Oscar Long-hin Lai. 2020. Gender
[17] Bernhard Ertl, Silke Luttenberger, and Manuela Paechter. 2017. The impact of stereotyping and STEM education: Girls’ empowerment through effective ICT
gender stereotypes on the self-concept of female students in STEM subjects with training in Hong Kong. Children and Youth Services Review 119 (2020), 105624.
an under-representation of females. Frontiers in psychology 8 (2017), 703. [45] Marie Travers, Jenna Bromell, Katie Crowley, James Vincent Patten, and Ita
[18] Camilla R Fisher, Christopher D Thompson, and Rowan H Brookes. 2020. Gender Richardson. 2023. Shooting for the Stars: How a STEM Initiative has Evolved to
differences in the Australian undergraduate STEM student experience: a system- Address Gender Challenges in Work and Education. In IEEE/ACM 4th Workshop
atic review. Higher Education Research & Development 39, 6 (2020), 1155–1168. on Gender Equity, Diversity, and Inclusion in Software Engineering (GEICSE). IEEE.
[19] Alicia García-Holgado, Amparo Camacho Díaz, and Francisco J García-Peñalvo. [46] Bianca Trinkenreich, Ricardo Britto, Marco A Gerosa, and Igor Steinmacher. 2022.
2019. Engaging women into STEM in Latin America: W-STEM project. In 7th Int. An empirical investigation on the challenges faced by women in the software
Conf. on Technological Ecosystems for Enhancing Multiculturality. 232–239. industry: A case study. In ACM/IEEE 44th Int. Conf. on Software Engineering:
[20] Susana González-Pérez, Ruth Mateos de Cabo, and Milagros Sáinz. 2020. Girls in Software Engineering in Society. 24–35.
STEM: Is it a female role-model thing? Frontiers in psychology 11 (2020), 2204. [47] Melvin Vooren, Carla Haelermans, Wim Groot, and Henriette Maassen van den
[21] Lucia Happe and Kai Marquardt. 2023. RockStartIT: Authentic and Inclusive Brink. 2022. Comparing success of female students to their male counterparts in
Interdisciplinary Software Engineering Courses. In 2023 IEEE/ACM 4th Workshop the STEM fields: an empirical analysis from enrollment until graduation using
on Gender Equity, Diversity, and Inclusion in Software Engineering (GEICSE). IEEE. longitudinal register data. Int. Journal of STEM Education 9, 1 (2022), 1–17.
[22] Catherine Hill, Christianne Corbett, and Andresse St Rose. 2010. Why so few? [48] Elijah Zolduoarrati and Sherlock A Licorish. 2021. On the value of encourag-
Women in science, technology, engineering, and mathematics. ERIC. ing gender tolerance and inclusiveness in software engineering communities.
[23] Sonja Hyrynsalmi and Sami Hyrynsalmi. 2019. Software engineering studies Information and Software Technology 139 (2021), 106667.
attractiveness for the highly educated women planning to change career in

36