Brazilian Girls’ Perspectives on STEM Careers

Abstract

This research examines the professional aspirations of teenage girls attending secondary school in a Brazilian state and how they perceive potential careers in Science, Technology, Engineering and Mathematics (STEM). STEM remains a masculinised field, not only due to the predominance of men but also because of enduring cultural representations that associate it with exceptional academic performance and highly technical competencies. In response, various educational intervention initiatives have sought to challenge these assumptions and promote the inclusion of girls and women in STEM. This article discusses an intervention carried out in four public schools as part of a project led by the Centre for Informatics at the Federal University of Paraíba. The project collected quantitative and qualitative data on subject preferences, professional expectations and perceptions of STEM among teenage girls. A mixed-methods exploratory approach was adopted to analyse the data and therefore to contribute to understand better the specific challenges of implementing such interventions in socioeconomically disadvantaged contexts, which remain underrepresented within the Women in STEM scholarship. The findings illustrate how social and financial constraints shape these young women’s career aspirations, intensifying both the desire for economic stability and the pursuit of personal fulfilment. At the same time, many perceive the future as uncertain and regard STEM as difficult to access, frequently associated with anxiety surrounding core science subjects. This article contributes to ongoing debates on gender and STEM education, offering insights into the Brazilian context and outlining considerations for the development and refinement of future school-based interventions.

1. Introduction

The 21st century is a technological era. One prevalent assumption is that Science, Technology, Engineering, and Mathematics (McDonald 2016; Sar 2021). At the same time, a key concern is with shortages of qualified STEM professionals (McDonald 2016; Sar 2021) as there has been a growing demand for STEM skills in the job market in recent years. White argues that this area has the potential to either reduce or reproduce social inequalities (White 2014). Among social inequalities, gender asymmetries in the job market are a mainstream social problem. European Union has a 30-year commitment to gender mainstreaming but still much remains to be done (Rubery et al. 2024). Some authors point out that in recent years, technology evolution consequences in terms of standard equality rights have been overlooked (Rubery et al. 2024). At the same time, STEM remains one of the fields with bigger differences between women and men participation. There is a persistent trend of women’s lack of presence and interest in careers requiring STEM-related skills (STEM Women 2023). This context has led to strategic investments in STEM education worldwide in recent decades. STEM education is expected to provide students with critical thinking skills for creative problem-solving (Cooper and Heaverlo 2013) and, ultimately, to make them more marketable in the workforce (White 2014). STEM global educational initiatives have focused mainly on increasing the number of women students pursuing STEM fields.

This article discusses STEM education drawing on a project that has been running continuously for over a decade. The project ‘Meninas na Computação’/Girls in Computer Science is based at the Informatics Centre of the Federal University of Paraíba (UFPB) in northeastern Brazil, and it is a university extension activity. Extension activities are initiatives developed by educational institutions with the aim of transferring the knowledge produced within these institutions to the broader community (Oliveira and Goulart 2015). These activities seek to foster interaction between academia and society, contributing to civic education and social development. It has the main objective of awakening the interest of high school students to the STEM area, more precisely Engineering and Computing, through debates, lectures and practical training (United Nations Academic Impact 2023).

The authors examine the most recent dataset from the project, seeking to understand current young women’s perspectives on a STEM career, as well as the specific features and challenges of such an intervention in a disadvantaged socio-economic context. Data analysis is conducted in dialogue with the existing literature. In view of the scarcity of research on this subject in the South American context (García-Holgado et al. 2019), the study aims to contribute to filling this gap and to giving greater visibility to the Brazilian reality. Following an introductory review of the relevant literature—addressing topics such as gender asymmetries in labour and science, particularly in STEM; the specificities and challenges of STEM education from a holistic perspective; and an overview of initiatives designed to tackle these issues—the article proceeds to present the intervention under analysis, detailing the Meninas na Computação project and the methodological approach adopted. Afterwards, the results and analysis are presented. The conclusions contribute to the body of knowledge on this theme by highlighting both the continuities and specificities of the case under study, while also reflecting on relevant improvements for the project and for STEM education more broadly.

1.1. Gender Asymmetries in Labour and Science as Resistant Societal Problems

Gender inequality in the labour market is a complex and multifaceted phenomenon that transcends geographical and historical boundaries, shaping women’s personal and professional trajectories throughout life. Until recently, the International Labour Organization (ILO) measured the gender pay gap solely on the basis of wages. Since 2022, however, it has adopted a broader labour income indicator that includes the earnings of all workers, acknowledging that a substantial proportion of the global workforce is self-employed—a group previously overlooked by wage-based measures. Under the former methodology, the global gender pay gap stood at 36.7 percent in 2018 (Eurostat 2019), whereas the new indicator reveals that women earn, on average, only 51 cents for every dollar of labour income earned by men (International Labour Organization 2023). If current trends persist, the ILO estimates that it will take around 70 years to close this gap. Complementing this measure, the recently introduced concept of the job gap (International Labour Organization 2024) provides further insight into gender disparities in access to employment, highlighting particularly severe disadvantages faced by women in low-income societies, who are often disproportionately affected (International Labour Organization 2024).

Gender inequality in the labour market is a complex and multifaceted phenomenon that transcends geographical and historical boundaries, shaping women’s personal and professional trajectories throughout life. Until recently, the International Labour Organization (ILO) measured the gender pay gap solely based on wages. Since 2022, however, it has adopted a broader labour income indicator that includes the earnings of all workers, acknowledging that a substantial proportion of the global workforce is self-employed—a group previously overlooked by wage-based measures. Under the former methodology, the global gender pay gap stood at 36.7 percent in 2018 (Eurostat 2019), whereas the new indicator reveals that women earn, on average, only 51 cents for every dollar of labour income earned by men (International Labour Organization 2023). If current trends persist, the ILO estimates that it will take around 70 years to close this gap. Complementing this measure, the recently introduced concept of the job gap (International Labour Organization 2024) provides further insight into gender disparities in access to employment, highlighting particularly severe disadvantages faced by women in low-income societies.

Contrary to the long-standing assumption of gender neutrality (Augusto et al. 2018), science is not exempt from these broader labour market inequalities (Penner 2015). As this paper will demonstrate, science—and STEM disciplines in particular—remain domains in which women face significant disadvantages (Eagly 2020). This inequality is evident not only in women’s underrepresentation across Science, Technology, Engineering and Mathematics, but also in its wider structural consequences, such as the perpetuation of gender gaps (Rosser and Taylor 2009) and the reduced quality and diversity of scientific output (Hammond et al. 2020).

Persistent gender disparities are visible in multiple dimensions of academic life. Huang et al. (2020) identified a long-term publication gap that consistently favours men, a trend exacerbated during the COVID-19 pandemic, when women’s research productivity declined amid increased domestic and caregiving responsibilities (Ryan et al. 2023; França et al. 2021). Inequalities also persist in academic leadership and research funding opportunities (Herschberg and Berger 2015), with both explicit and implicit biases in merit evaluations undermining women’s career progression (Llorens et al. 2021).

These patterns reinforce an academic culture that often undervalues women’s contributions. STEM fields are particularly affected, as they continue to be perceived as ‘male domains’, perpetuating gender stereotypes (Belo and Estébanez 2022). The uneven representation of women in disciplines such as Technology, Engineering and Mathematics not only limits women’s professional advancement but also constrains scientific progress.

At its essence, science aspires to objectivity, innovation, and the advancement of knowledge. Yet research consistently shows that diversity fosters creativity, enhances problem-solving, and strengthens scientific performance (Love et al. 2022). The persistent underrepresentation of women in senior and prestigious positions restricts the plurality of perspectives essential to innovation. Even when women enter these fields, they continue to encounter barriers to recognition and creativity, often being perceived as less original than their male peers (Kronqvist and Rossie 2024). Consequently, women and other marginalised groups remain unable to contribute fully to scientific advancement. The lack of gender equity in science, therefore, constitutes not only a matter of social justice but also a structural impediment to innovation.

Although girls perform as well as or better than boys in international assessments of scientific literacy (Stoet and Geary 2015), women continue to earn fewer university degrees in STEM fields across all countries, suggesting a significant loss of female talent between secondary and higher education (Aires Moreira and Mattos 2018; Stoet and Geary 2018). Given the global demand for qualified STEM professionals, this persistent disparity underscores the urgency of initiatives aimed at promoting gender equality in STEM education and careers (Sungur Gul et al. 2023; Pal et al. 2024).

1.2. Identified Challenges in STEM Education

In OECD countries, women now pursue higher education at higher rates than men. On average, 52% of young women aged 25 to 34 hold a university degree, compared with 39% of men (OECD 2021a). This gap in favour of women widened between 2010 and 2020. However, this overall advantage masks pronounced disparities across fields of study: while women dominate in areas such as health and welfare, they remain markedly underrepresented in STEM disciplines (OECD 2021a). The Global Gender Gap Report (Pal et al. 2024) further illustrates this imbalance, showing that women account for just 28.2% of the global STEM workforce, despite representing nearly half (47.3%) of the non-STEM workforce (World Economic Forum 2024). The STEM education approach emerged as a response to the growing need to enhance students’ competitiveness, critical thinking, and motivation to pursue scientific careers. Its goal is not only to develop technical and analytical skills but also to prepare learners to address current and future societal challenges of a scientific and technological nature. As Bonito and Oliveira (2022) note, STEM education represents a complex and ongoing global endeavour that has drawn increasing academic attention over recent decades.

Recent data suggest that these efforts are beginning to yield results. The number of STEM graduates worldwide continues to rise, led by Asian countries such as China and India, which recorded the highest figures in 2020 (Oliss et al. 2023). Brazil, the focus of this study, ranked sixth globally, with a notable 26% increase in STEM graduates between 2015 and 2020. However, according to the 2023 Brazilian Higher Education Census (INEP 2024), women—who represent around 60% of all undergraduate degree recipients—account for only 36.64% of STEM graduates. This shows that, despite progress, gender disparities persist. OECD (2021b) data reinforce this imbalance: between 2016 and 2019, only 14.25% of girls across 44 countries enrolled in engineering programmes, with Brazil ranking 24th (12.83%).

Explanations for these persistent gaps are multifaceted. A systematic review by Beróíza-Valenzuela and Salas-Guzmán (2024) identifies the combined influence of psychological, contextual, and sociocultural factors on students’ choices and performance in male-dominated fields. Context is a crucial element in educational research (Berliner 2002): although women today receive more encouragement than in the past, their experiences often reflect imbalances, bias, and limited institutional support (Blackburn 2017). Female faculty members frequently report a lack of mentorship and work–life balance policies, which contributes to heightened stress and hampers progression.

Teachers play a pivotal role in this process. As key mediators of learning experiences, they can identify both the advantages and the challenges of STEM teaching. According to Sungur Gul et al. (2023), the most frequently cited difficulties are the highly technical nature of the content and the significant preparation time required. Yet the same authors highlight a scarcity of studies from South America, underlining the need for greater research attention to this context. Broader analyses indicate that while STEM education research is expanding globally, cross-country comparisons remain premature due to the limited number of studies and contextual diversity (Sar 2021; Ardies et al. 2015; Beróíza-Valenzuela and Salas-Guzmán 2024).

A holistic understanding of gender disparities in STEM requires consideration of interrelated factors such as stereotypes, self-efficacy, expectations, and motivation. From an early age, individuals internalise gendered representations of science and scientists, often perceiving subjects like mathematics and physics as masculine domains (Archer et al. 2010; Fine 2018; Hill et al. 2010; Makarova et al. 2019). These perceptions negatively affect career aspirations, particularly among girls. The scarcity of female role models compounds this effect, reinforcing exclusionary cycles (Hill et al. 2010; Milgram 2011). Teachers and parents are among the most influential role models: having a teacher or parent working in STEM significantly increases students’ aspirations towards these careers (Ardies et al. 2015; European Commission et al. 2021).

The internal culture of STEM education also plays a decisive role. Stereotypes about competitiveness and the absence of collaboration in STEM professions contribute to women’s perception that these fields are “not for them” (Hill et al. 2010; Diekman et al. 2010; Zachmann 2018). A study by the organisation Força Meninas (2023) in Brazilian schools illustrates how socioeconomic and cultural contexts shape career expectations: girls in public schools tended to imagine futures in professions associated with authority and protection, while those in private schools preferred medicine and business—neither group identifying STEM careers as desirable. Notably, 57.1% of surveyed children reported not knowing any woman working in STEM, underscoring the absence of visible role models.

Motivation and interest are closely linked to self-efficacy—the belief in one’s ability to succeed in specific tasks—which influences persistence and engagement in STEM learning (Bandura 1982; Rundgren et al. 2019; Chan 2022). These beliefs develop through learning experiences that are themselves embedded in broader cultural and social contexts (Beróíza-Valenzuela and Salas-Guzmán 2024). Research shows that although girls often achieve equal or higher grades than boys, they tend to underperform in applied problem-solving and transfer tasks (Hyde et al. 1990; Mendes-Barnett and Ercikan 2006). Inclusive pedagogies, critical thinking, and experimental approaches can mitigate these challenges (Suhirman and Prayogi 2023).

Nonetheless, many young women continue to perceive STEM as unwelcoming or irrelevant to their aspirations (Brophy et al. 2008; Chan and Cheung 2018). The “leaky pipeline” framework highlights how differences in self-efficacy, interests, and perceived value contribute to attrition across educational stages. This tendency is reinforced by the general decline in STEM interest during secondary education, which makes sustained, engaging learning experiences essential for developing “scientific capital” (European Commission et al. 2021; Beróíza-Valenzuela and Salas-Guzmán 2024). Ultimately, the combination of self-efficacy and outcome expectations shapes students’ interest in STEM: when girls underestimate their abilities, socially constructed boundaries often prevent them from pursuing engineering or related careers, even when performance is high.

Despite extensive research identifying multiple factors shaping women’s trajectories in STEM, the gender gap remains largely unresolved (Sar 2021). At the same time, the geographical coverage of this phenomenon remains uneven: empirical evidence from South American contexts is still scarce, and there is a clear lack of mixed-method data on girls’ STEM aspirations in the region. This highlights the urgent need to evaluate and strengthen the effectiveness of existing interventions across South America. It is within this context that the next section turns to the dimension of intervention.

1.3. What Is Being Done? The Initiatives Promoting STEM Among Girls and Women

Faced with the under-representation of girls and women in STEM, various institutions and groups have been developing projects to change this scenario. In general, these projects seek to demystify access to STEM and provide role models by involving women scientists and technology leaders (Moreira et al. 2020). Participatory and student-centred methodologies are often used (Santos et al. 2019). Interventions also make resource of peer education, involving young women university students in STEM (Peres 2021) as an intermediate form of role modelling and which aims to deconstruct the prevalent idea that it is very difficult to pursue university studies in STEM. Women NGOs play an important role in these projects (Guthridge et al. 2022) that in recent decades have been prioritised by international organisations such as the United Nations or UNESCO. In Europe, there are specific funding lines for the inclusion of women in STEM within the broader funding line for promoting gender equality (Sabourin and Jones 2023).

These programmes not only introduce girls to more practical approaches to STEM, but also provide emotional support and, for older students, opportunities to establish contact with professional networks, thereby strengthening their social capital (Bourdieu 1984). At South America, where the educational challenges are complex (Gentili 2009), these projects have the added value of bringing young people resources that they often do not have the opportunity to contact because access to technology is limited in schools (Salvatierra and Kelly 2023).

Turning to the Brazilian context, Costa et al. (2020) identified 25 initiatives aimed at promoting gender equality in STEM education, as shown in the following

Table 1. STEM initiatives for gender equality in Brazil.

Name of the Initiative	Institution	Instrument Type (1)	STI Gender Objectives	Beneficiaries (2)	Geographical Coverage	SDG (3)
Ciência para Meninas	Extension project—State University of Piauí	C-E	1 2 3 4 7	P-R	State	4-5
Cloud Girls	Community supported by several large companies	C-E	1 3 4 7	P-Q-R	National	4-5
Debian/Women	Debian-supported community	C-E	1 3 4 7	P-Q-R	International	5
Django Girls	Django-supported community	C-E	1 3 4 7	P-Q-R	National	5
Elas na Engenharia	Extension project—University of São Paulo	C-E	1 2 3 4 5 7	P-Q-R	State	4-5
Girls Power Tech	CISCO	C-E	1 2 3 4 5 7	P-Q-R	International	4-5
MariaLab	MariaLab	C-E	1 2 3 4 5 7	P-Q-R	National	5
Meninas Digitais	Brazilian Computer Society	B-C-E	1 2 3 4 5 6 7	P-Q-R	National	4-5
Meninas na Ciência	Physics Institute—Federal University of Rio Grande do Sul	C-E	1 2 3 4 5 6 7	P-Q-R	State	4-5
Minerv@s Digitais	Extension project—University of Rio de Janeiro	C-E	1 2 3 4 5 6 7	P-Q-R	State	4-5
MinervaRockets	Technology College—UFRJ	C-E-D	1 2 3 4 5 6 7	A-B-C-D-N-P-Q-R	State	4-5
MM360	L’Oréal Brazil with support from UNESCO	G-B	1 3 4 5 6 7	P-Q-R-H-M	National	5-8
Para Mulheres na Ciência	L’Oréal Brazil with support from UNESCO	G-B	1 3 4 5 6 7	P-Q-R	National	5-8
Pretalab	Olab supported by The Ford Foundation	C-E	1 3 4 5 6 7	P-Q-R-U-S	National	5-8
Programaria	Intel-supported community	C-E	1 3 4 5 7	P-Q-R	National	4-5
PyLadies Manaus	Python Software Foundation	C-E	1 2 3 4 5 7	B-C-N-P-Q-R	State	4-5
PyLadies Paraíba	Python Software Foundation	C-E	1 2 3 4 5 7	B-C-P-Q-R	State	4-5
Reprograma	Community supported by several large companies	C-E	1 3 4 5 6 7	P-Q-R	National	4-5
Simpósio Brasileiro Mulheres em STEM	Governments	C-E	1 2 3 4 5 6 7	P-Q-R	National	5-8
STEMPD	ITA supported by Johnson & Johnson	C-E	1 2 3 4 5 6 7	P-Q-R	National	5-8
WoMakersCode	Supported by Google and Microsoft	C-E	1 3 4 5 7	P-Q-R	National	4-5
Women Entrepreneurship	Fund supported by L’Oréal	A-C-E	1 3 4 5 7	P-Q-R-J-H-M	International	5-8
Women in Engineering—WIE	IEEE	C-E	1 3 4 5 7	P-Q-R	International	4-5
Women Techmakers	Google	C-E	1 3 4 5 7	P-Q-R	International	4-5

(1) Instrument type corresponds to the following categories: A—Technical Assistance; B—Scholarships/Fellowships; C—Training; D—Awards and Competitions; E—Creation and aid of technological poles, excellency centres, and communities; F—Donations (individuals/companies); G—Fairs; H—Trust; I—Financial Guaranty; J—Credit incentives and venture capital; K—Fiscal incentives; L—Loans; M—Information Services; N—Subsidy (non-reimbursable contributions). (2) “Beneficiaries” corresponds to the following categories: A—Research centres; B—Universities; C—Schools/Colleges/Institutes; D—Technical training centres; E—Public institutes; F—Professional institutes; G—STI public or private non-profit organizations; H—Private companies; I—Small and medium-sized companies; J—Cooperatives; K—Foundations; L—Local R&D groups; M—Ad hoc associations; N—University lecturers and researchers; O—Technical staff and assistants in STI; P—Students; Q—Individuals; R—Women (exclusively); S—Indigenous peoples and local communities; T—Disabled people; U—Minorities; V—Professionals/Ph.D.s. (3) “Sustainable Development Goal (SDG)” corresponds to the following goals: 4—Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all; 5—Achieve gender equality and empower all women and girls; 8—Promote sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all. Source: (Costa et al. 2020).

, reproduced from the authors.

The majority of these initiatives were promoted by companies and universities, with aims such as ‘attracting girls to science and technology careers; fostering interest in the exact and technological areas; encouraging and supporting the participation of women in the technological area; encouraging more girls and women to learn about programming; helping more women to become active in free software communities; and encouraging girls to pursue careers in STEM’ (Costa et al. 2020, p. 1254).

In the last decade, CNPq (National Council for Scientific and Technological Development) has launched public calls specifically aimed at promoting the inclusion of women in technology. The first call, 18/2013 MCTI/CNPq/SPMPR/Petrobras—Girls and Young People in Exact Sciences, Engineering and Computing, was issued in 2013 and approved 325 projects across Brazil, including the Meninas na Computação project (CNPq 2013). A second edition, CNPq/MCTIC Nº 31/2018—Girls in the Exact Sciences, Engineering and Computing, was launched in 2018 but approved only 70 projects nationwide, marking a substantial reduction in investment and leaving out many regions and schools (CNPq 2018). More recently, a UNESCO (2022) mapping of initiatives to encourage girls and young people in STEM identified 217 projects in the country.

According to Allueva-Pinilla et al. (2019), the advantage of involving universities in such projects is that they ‘not only promote the role of women in STEM areas, but also bring middle school students closer to the university and foster the enrolment of students, especially women, in the first courses of scientific and technical education’ (p. 1348).

Despite the notable number of initiatives, knowledge about the Brazilian reality remains scarce, with important exceptions of works like Costa et al. (2020, 2024) and Oliveira et al. (2019). There is little understanding of how these initiatives were developed, including their potentialities and shortcomings. This study seeks to contribute towards filling this gap in the literature.

2. Materials and Methods

This article presents the more recent outcomes of the Meninas nas Computação project between 2019 and 2023. The project is part of the national initiative Meninas Digitais launched by Brazilian Computer Society (line 8 in

Table 2. Ten Major Areas of Educational Programmes in Brazil.

01 Education

02 Arts and Humanities

03 Social Sciences, Journalism and Information

04 Business, Administration and Law

05 Natural Sciences, Mathematics and Statistics

06 Information and Communication Technologies

07 Engineering, Manufacturing and Construction

08 Agriculture, Forestry, Fisheries and Veterinary

09 Health and Welfare

10 Services

Source: (INEP 2019).

). Launched in 2014 and funded by the first CNPq call, the Meninas na Computaçáo project represented a pioneering initiative in promoting STEM engagement among women students in the state of Paraíba. Until now the project has reached over 2000 students across various schools and regions within the state.

The project’s overarching objective is to inspire and cultivate young girls’ interest in STEM—particularly in Computer Science—through training in Information and Communication Technologies (ICTs). It seeks not only to develop technical skills and foster vocational interest in computing, but also to empower participants by enhancing their sense of self-efficacy and preparing them for future enrolment in computing programs at either the technical or higher education levels. Computer science is one of the STEM areas were women less represented in Brazil, around 11%, a value well below the already low participation of girls in STEM education (Costa et al. 2024).

The Meninas na Computação initiative targets students from public secondary schools, covering all three grade levels and primarily reaching teenagers aged 14 to 17. Until 2017, schools were selected from a list provided by the funding agency (CNPq 2013; CNPq 2018). Since then, with the project continuing based on volunteer work, schools have been chosen according to their expressed interest, the availability of a computer lab, and the presence of a teacher able to act as supervisor—conditions considered essential for ensuring the sustainability of the intervention. Proximity to the university campus is also considered, given the logistical needs of the project team, especially of the BA students. According to the organisers, several schools have been excluded due to these constraints.

The implementation team consists of faculty members from Centre for Computer Science, together with undergraduate students from its three programs: Computer Engineering, Computer Science, and Data Science and Artificial Intelligence. The involvement of undergraduate women in STEM is a central pillar of the project, intended to create opportunities for informal conversations and peer modelling (King et al. 2021). However, in recent years this has become more difficult, as no funding has been available to provide grants, unlike in the project’s early stages.

The activities developed within the project combine technical training, socio-cultural reflection, and hands-on practice. It is organized through a set of interconnected methodological components aimed at engaging students with issues of gender, technology, and education. It combined thematic lectures and debates, peer dialogue, practical workshops, and technical visits, each designed to address different dimensions of participation in computing. The lectures, focused on gender inequalities in science and technology with a particular emphasis on women’s trajectories, were not conceived merely as knowledge-transfer sessions. By being followed by structured debates, they created spaces where participants could share personal experiences of discrimination and differential treatment, allowing these accounts to become part of a collective process of recognition and critical reflection.

Complementing this dimension, undergraduate students from computing programs at UFPB interacted with high school participants through dialogues that centred on navigating male-dominated academic environments, overcoming barriers, and facing the challenges of learning programming. These exchanges functioned as informal or near-peer mentoring moments, offering models of possible academic trajectories while also exposing the persistence of structural obstacles to women’s participation in computing.

Alongside these conversations, technology training workshops provided participants with hands-on experience in programming and digital creation. Activities included introductory programming with Scratch, mobile app development through MIT App Inventor, game design with Construct 2, and blog creation on topics chosen by the students themselves. By adopting accessible and playful platforms, the workshops sought to reduce entry barriers to technology while encouraging students to see digital tools as resources for creativity and self-expression rather than purely technical skills.

The methodology also incorporated guided visits to laboratories and academic programmes at UFPB, particularly those dedicated to Artificial Intelligence, Robotics, Drones, and Software Development. These visits placed participants in direct contact with advanced research environments and were designed to expand their horizons regarding higher education in computing. Experiencing these spaces first-hand was intended to foster a tangible sense of belonging in technological fields and to reinforce the idea that such environments can be accessible to them.

The data analysed in this article were collected between 2019 and 2023 and must therefore be situated within the broader contextual constraints of the COVID-19 pandemic, which was felt both in the structure of the intervention and in the number of participants. At the outset of the project in 2019, 162 students participated. This number dropped significantly to just 39 in 2020, but gradually recovered in the following years, increasing to 50 students in 2021, 92 in 2022, and 130 in 2023. The project relies on in-person visits to schools to establish initial contact, followed by remote classes. Therefore, the inability to carry out face-to-face interactions had a considerable impact on participation rates. Only in 2023 did the number of participants approach the level observed before the pandemics.

In what concerns the structure, during the years of 2020 and 2021 the project went through pandemic confinement restrictions. In Brazil it lasted long so the team adapted the project to a hybrid structure and only schools with adequate technological infrastructures were able to participate. This is not the reality for a considerable part of Paraiba schools, especially in the inner state.

The success of the project is regularly assessed through participants’ satisfaction with their learning experience, using a Likert scale ranging from 1 (Poor) to 5 (Excellent). For the dataset here analysed, a total of 263 responses were collected, with 53% rated as Excellent, 40% as Good, 5% as Neutral, and only 2% as Fair or Poor. In other words, 93% of the students evaluated the learning experience positively. This is an encouraging result, especially considering it was their first exposure to programming in their academic journey (Mattos et al. 2023).

The Meninas na Computação project applies a research protocol as initial assessment of the target population of the intervention. This protocol consists of a semi-structured questionnaire comprising 25 questions, both closed-ended and open-ended. The instrument was developed by project team, drawing on previously validated tools (Grings et al. 2018; Pinto et al. 2017). The structure of the questionnaire was divided into four main sections:

Part 1: School identification, gender, age, and type of previous schooling (public or private).

Part 2: Preferences regarding school subjects and the use of technology in educational settings.

Part 3: Intentions concerning future academic courses and career aspirations after completing secondary education. Open-ended prompts included questions such as the students’ desired programme after completing high school and, “How do you imagine yourself 15 or 20 years from now?”

Methodological Approach

The research presented here draws on empirical data collected during the implementation period from 2019 to 2023. During this time, the project was carried out in four state secondary schools (see

Table 3. Number of Students by Participating School.

School Name	Number of Students
Escola Estadual Rebeca Simões	274
Escola Estadual João Pereira Gomes Filho	119
Escola Estadual Celestin Malzac	49
Escola Estadual José Rocha Sobrinho	30
Total	472

Source: Authors’ elaboration.

), yielding a total of 472 valid responses.

This study aims to address the knowledge gaps identified above by examining the implementation of an intervention initiative in the Brazilian context and exploring the participating girls’ perspectives on pursuing a career in STEM.

The methodological approach adopted for this study is best described as a mixed-methods design, focusing on a selected subset of variables from the overall dataset. This deliberate selection allowed for a balanced and sequential consideration of both closed- and open-ended responses (Ahmed et al. 2024).

While the study integrates both quantitative and qualitative data, the qualitative analysis of open-ended variables received particular emphasis. This methodological choice is justified by the exploratory nature of career aspirations and their underlying social and personal factors, where a detailed understanding of the participants’ own words is particularly important (Hawkins 2017). Descriptive statistics alone are insufficient to fully capture the complexity and nuance of the participants’ motivations. This heightened explanatory focus was therefore essential for the comprehensive triangulation of data and the robust interpretation of the quantitative findings.

Regarding the researchers’ positionality and ethical considerations, one of the authors is actively involved in the Meninas na Computação project, while the other two are entirely external to it. The ongoing dialogue and collaborative teamwork helped to mitigate—though not eliminate—possible preconceptions or proximity-related bias. A particularly effective aspect of the process was that the project coordinator had to ‘tell the story’ of the initiative to the rest of the team and respond to their critical questioning. This external perspective generated a proliferation of questions that might not have emerged otherwise, encouraging deeper reflection on the research process, as advocated by Creswell (2014). Moreover, the team actively drew on its cultural diversity, treating it as an asset that further enriched the analysis.

Regarding the quantitative approach, the authors employed univariate and bivariate descriptive statistics, as well as tests of association—specifically, the chi-square test. All statistical procedures were performed using SPSS software, version 30.

Regarding the procedures applied to the open-ended variable on the intended study programme, the responses ‘Don’t know’ (f = 31), ‘No answer’ (f = 5), and ‘CFO’ (f = 51) were disregarded. The ‘Don’t know’ responses (f = 31), however, were subjected to a dedicated analysis. ‘CFO’ refers to the Officers’ Training Course; since Rebeca Simões School is a military institution, there is a strong tendency for students to choose this pathway. Although relevant, as highlighted in the literature (Força Meninas 2023), the authors have considered it a bias in the context of this research and therefore excluded those responses. This process resulted in 385 valid answers, which were subsequently categorised into the educational areas defined by UNESCO (2014) for the Brazilian context.

The International Standard Classification of Education (ISCED) of UNESCO serves as a comprehensive and internationally recognized framework for the systematic organization of educational programmes and qualifications. By applying standardized and universally agreed definitions, ISCED facilitates meaningful comparisons of education systems across countries. It plays a vital role in ensuring statistical consistency, supporting cross-national research, and informing evidence-based policymaking in the field of education (UNESCO 2014). Brazil uses ISCED adapted to the Brazilian educational reality (INEP 2019).

According to this classification, the ten major areas are:

The two open-ended response variables were qualitatively analysed MAXQDA version 2024.

The first is the variable Reason for choosing the intended programme and it was analysed by performing categorial content analysis. Several levels of categorical refinement were tested, based on the reviewed literature and through comparative analysis with the second variable. In the end the responses were grouped into the following refined categories: Dream/Vocation, Job Market/Financial Reasons, Personal Interest/Affinity, Family Influence, School/Teacher Influence, Influence of Professionals/Role Models, Personal Choice/Autonomy, No Specific Motivation, and Other.

This variable was also examined using a word cloud generated with the software wordclouds.com. Although exploratory in nature, this technique represents a form of quantitative text analysis, as it counts the frequency of words in the dataset. It is particularly suitable for processing open-ended survey responses or large volumes of textual information (Vilela et al. 2020), allowing the most frequently occurring words to be visually highlighted and thereby revealing predominant patterns. Nonetheless is essential to exercise great caution when handling stop words, as improper treatment can easily distort the results (Heimerl et al. 2014).

The second variable is the last question of the survey, formulated as Where do you imagine yourself in 15-20 years? This short-answer, open-ended question was subjected to a structural content analysis, maintaining a taxonomic level (Amado et al. 2014). The category system was developed using a mixed process. After a literature review, the authors chose to categorize our observations using the term ‘spheres of life’. This term refers to the various domains where individuals engage in daily life and fulfil their social roles. These spheres as intrinsically linked with Social Role Theory (Eagly and Wood 2012) because the expectations tied to specific roles often vary significantly from one life sphere to another. The categories and subcategories used to classify the constituted corpus are: Personal/Individual Sphere, Social and Leisure Sphere, Family Sphere and Professional/Financial Sphere.

3. Results and Discussion

This section presents the sociodemographic profile of the study participants, offering a contextual basis for interpreting their academic choices and career aspirations.

A histogram of age distribution revealed a centred but slightly asymmetric pattern, with a predominance of students aged 15 and 16, which is consistent with expectations for this group.

The next table shows the number of participants per school., where the two first schools have more than 80% of the participants.

The students surveyed are concentrated in the first year of high school (Figure 1).

Most of the study participants attended both public and private schools, with only 29% having studied exclusively in the public system. This is of great importance in the Brazilian context, which is marked by a sharp disparity in the quality of public and private education at the level of basic schooling (Guimarães and Sampaio 2009).

The result concerning the participants’ intended programme wordclouding is presented in Figure 2 below. After removing generic terms, the graph shows the most frequently mentioned courses by students, providing an overall view of their academic preferences.

The area with the highest concentration of choices is Health and Welfare, a trend that aligns with the persistent predominance of women in this field globally. This pattern is well-documented in the academic literature (Steiner-Hofbauer et al. 2023; Scheffer and Cassenote 2013).

Within this broad area, 28% of the students indicated a specific interest in pursuing a degree in Medicine. In the Brazilian context, Medicine is one of the most competitive undergraduate programmes, consistently ranking first in admission demand at public, tuition-free universities. Conversely, in private higher education institutions, the cost of a medical degree is prohibitively high, ranging from six to twelve times the national minimum wage, depending on the region (Andrade 2025).

Students’ career intentions were grouped into STEM and non-STEM categories, following the standard classification framework (INEP 2019). As shown in

Table 4. Distribution of Career Intentions by STEM and Non-STEM Areas.

Category	Frequency	Percentage
Non-STEM	325	84.4%
STEM	60	15.6%
Total	385	100%

Source: Authors’ elaboration.

, prior to the intervention most students (84.4%) expressed interest in non-STEM fields, while only 15.6% indicated preferences aligned with STEM. When cross-tabulated with students who had always attended public schools, this percentage dropped to just 6.5%. These results reinforce findings in the literature, which consistently highlight the low level of interest in STEM careers among (young) women (OECD 2021a; World Economic Forum 2024). Moreover, Medicine emerged as the primary aspiration among those opting for non-STEM pathways, in line with research showing women’s continued overrepresentation in health-related disciplines (Madriaga et al. 2022; OECD 2021b; INEP 2024). We now compare our data with key factors identified in the literature as contributing to students’ lack of motivation toward STEM.

The variable More Difficult Subject revealed a clear concentration in a “fearsome trio” of disciplines—Physics, Mathematics, and Chemistry—which stand out markedly from all others (see

Table 5. More Difficult Subject.

Subject	Mentions	Percentage (%)
Physics	324	22.27
Mathematics	301	20.69
Chemistry	272	18.69
History	76	5.22
Philosophy	73	5.02
Portuguese Language	73	5.02
Laboratory (Science Labs)	62	4.26
Geography	61	4.19
Foreign Language	60	4.12
Biology	52	3.57
Literature	39	2.68
Sociology	33	2.27
Physical Education	13	0.89
Arts	7	0.48
Programming	5	0.34
Programming Logic	3	0.21
Digital Games	1	0.07
Total	1455	99.99

Source: Authors’ elaboration.

). This finding is consistent with previous studies identifying science- and math-related anxiety as a key barrier to students’ engagement with STEM (Tomperi et al. 2020). However, our results add nuance to the existing literature: whereas prior research often highlights Mathematics as the primary source of apprehension (Wang 2013; Tyler-Wood et al. 2018), Physics emerged here as the subject eliciting the strongest negative reactions, followed closely by Mathematics and Chemistry.

This inversion suggests that apprehension toward STEM subjects is not uniform and may vary according to disciplinary perceptions and learning experiences. The cumulative effect of these negative associations likely acts as a deterrent to students’ pursuit of STEM-related academic and career paths.

While the quantitative results presented above provide an overview of students’ academic intentions and the general distribution of interest across fields, they do not fully capture the underlying motivations behind these choices. To gain a deeper understanding of the factors influencing students’ decisions, we conducted a qualitative analysis of the two open-ended response variables.

Beginning with the Reason for choosing the intended programme, there were 472 responses in total, of which 27 were non-responses (empty cells). These correspond to approximately 5.68% of missing data—a rate generally considered acceptable in many studies, although it warrants attention depending on the type and sensitivity of the analysis. After grouping responses by similarity and excluding CFO due to its specificity, eight categories of reasons were identified, as shown in Figure 3 below.

It is noteworthy that 33.5% of the participants reported having no specific motivation for what constitutes a major life choice. While this finding aligns with recent studies indicating increasing uncertainty among young people regarding their future (OECD 2024; Guo 2025), other possible explanations may also account for this trend. One immediate hypothesis is the impact of the pandemic context, which has heightened uncertainty and undermined students’ motivation, as highlighted in several studies (Cayubit 2024; Ismail et al. 2025). This finding points to the need for further investigation, which would greatly benefit from longitudinal research capable of capturing how these patterns evolve over time.

Family influence is a well-established factor in career aspirations (Workman 2015), underscoring the importance of social context. From a Bourdieusian perspective applied to the Brazilian context, Berlato et al. (2021) argue that family and social class reproduction exert greater influence than schooling. Terruggi et al. (2019) results also point out in the same direction.

A chi-square test was conducted to assess the association between students’ motivation categories and their first interests’ programme grouped by areas. A statistically significant association was found between motivation categories and programme areas (χ² = 147.45, p < 0.001), meaning that students’ motivations differ according to their intended field of study. This variation is displayed in Figure 4.

Motivational patterns are not homogeneous across fields. Health and Humanities are predominantly associated with intrinsic factors, expressed in responses such as ‘It has always been my dream’ and ‘Ever since I was a child, I have known that I wanted to be a doctor’. while Business and Law are more strongly linked to extrinsic motivations such as financial return or social prestige. Within Health-related areas, particularly Medicine, personal motivation emerged as the most frequent driver—a pattern consistent with findings reported by Stowers et al. (2019) and, to a lesser extent, Goel et al. (2018).

The representation of STEM fields in students’ motivational profiles was comparatively low. Among those expressing interest in STEM, motivations were mostly personal or family-related, though their frequency was marginal. Rather than simply confirming the low overall interest in STEM careers observed in the literature, these results highlight how STEM aspirations occupy a narrow motivational space within students’ broader academic choices. This finding contributes an additional layer of understanding by showing that even when girls express curiosity about STEM, such interest is often individualised and unsupported by wider institutional or social reinforcement.

The final variable analysed, My future in 15–20 years, captures how students envision their long-term goals. These expectations serve as an indicator of their likely choices and persistence in STEM careers and may also shed light on the reasons why some students decide not to pursue these fields. Of the 465 young women who shared their future expectations, most described aspirations centred on professional achievement, economic stability, and personal fulfilment. Some interesting examples of emblematic responses in each of these areas include: ‘to have a good job’ or ‘to be recognised in my profession’; ‘I want to be rich’; ‘to have my own house and car’ regarding financial stability; ‘to get married’; and ‘to be happy’ in relation to personal fulfilment. The formulation patterns of these answers suggest a strong alignment with dominant cultural values; however, further data would be required to explore this aspect in greater depth.

Career success and financial independence clearly emerged as the main priorities, reflecting a strong awareness of the link between education, autonomy, and social mobility in the Brazilian context. These findings add nuance to existing research by showing that girls’ aspirations are not limited by a lack of ambition or interest in scientific fields, but rather by perceptions of misalignment between STEM careers and the forms of fulfilment they value most—particularly stability, purpose, and work–life balance.

While previous studies have noted the influence of gendered stereotypes and the scarcity of female role models in deterring women from STEM (e.g., Belo and Estébanez 2022; Hill et al. 2010), our data suggest that these structural and cultural barriers operate through a more complex motivational pathway. For many participants, STEM professions appear incompatible with the kind of meaningful and sustainable careers they envision for themselves. In this sense, the underrepresentation of women in STEM reflects not only persistent stereotypes but also the limited visibility of career models that reconcile technical excellence with social contribution and personal wellbeing.

4. Conclusions

The gender gap in STEM fields remains a global concern, and STEM education plays a fundamental role in promoting inclusion. Many valuable initiatives have been developed, yet it is crucial to take stock of their progress and critically evaluate their strengths and weaknesses from a contextual perspective. More broadly, advancing gender equality in STEM requires deconstructing stereotypes, increasing the visibility of female role models, and reshaping social and family structures so that women are not forced to choose between career and personal life (IBGE 2022).

The study of the Meninas na Computação project in Brazil reflects international trends while also revealing specific cultural nuances. At the Federal University of Paraíba (UFPB), where the project operates, women’s participation remains below 30%. Nevertheless, since 2021, the proportion of female entrants in Computer Science programmes has been steadily rising (UFPB 2025). Although it is not possible to determine whether these students directly participated in the project, testimonials suggest that the initiative has positively influenced enrolment decisions.

Schools’ interest, students’ engagement, and their satisfaction with the learning experience are promising signs but insufficient on their own. A major obstacle—particularly in public schools—is the lack of infrastructure and adequately trained teachers for STEM subjects, a challenge that requires long-term investment. Continued mentoring is therefore essential, especially for first-year students who still have time to make informed academic and career decisions. More importantly, interventions must be critically assessed to understand how and why they work, moving beyond simple satisfaction measures.

In line with Straza (2024), interventions should begin early in schooling, as stereotypes about STEM professions emerge in childhood. Extracurricular projects need to accompany students throughout their education, complemented by structural change within formal education. Rapid technological and social transformations make this imperative, particularly for girls. Current choices in STEM will determine whether future societies move toward greater gender equality—or reproduce existing inequalities.

The literature shows growing attention to women’s participation in STEM, but research must move beyond workplace outcomes to examine the subjective and socio-cultural dynamics that shape career decisions. Factors such as family care responsibilities, domestic work, and the socialisation of girls into caregiving roles remain highly relevant in Brazil, where millions of adolescents leave school each year to support their families (INEP 2024; IBGE 2022). Effective interventions must therefore adopt a holistic perspective that links schools, families, and the media.

The findings of this study reinforce the importance of these subjective dimensions. Gender stereotypes—such as perceiving Physics, Mathematics, and Chemistry as difficult or masculine—remain significant barriers. Consistent with this, students in our sample identified these three subjects as their greatest academic challenges, even though the study did not explicitly address stereotypes.

Family influence also emerged as a key factor. While traditional gender roles persist in the girls’ future aspirations, family support strongly shaped their educational and career intentions. This finding reflects Brazil’s broader socio-cultural context, marked by patriarchy and inequality (Chein et al. 2007; Macedo et al. 2021). The tension between economic necessity and personal fulfilment further complicates decision-making. The desire for financial stability may lead young women to choose less demanding or more secure career paths than those typically found in STEM.

Motherhood also appears as a relevant though secondary consideration. The contrast between those who prioritise traditional family roles and those who seek fulfilment through work echoes broader global patterns but is particularly acute in the Global South (Santos 2014). Understanding how these life choices shape professional aspirations—especially in STEM—remains an important area for further research.

Finally, motivation for STEM appears deeply intertwined with gender. In this study, the emphasis young women placed on economic security suggests a powerful, underexplored lever for engagement. Interventions that align STEM participation with aspirations for financial independence could prove particularly effective.

Promoting genuine equity in STEM thus requires policies that not only attract women but also sustain their participation and success. These efforts must challenge stereotypes while addressing local socio-cultural realities. The persistence of gendered messages that glorify traditional femininity (Simili and de Souza 2015) underscores how urgent it is to create environments where women can see themselves—and be seen—as scientists, engineers, and innovators.

Study Limitations and Suggestions for Future Research

This exploratory study set out to examine the underrepresentation of women in STEM, aiming to generate insights and formulate questions for future research. However, its conclusions must be interpreted with caution, given several methodological and contextual limitations.

The sample cannot be considered representative of all state schools in Brazil, or even within the state of Paraíba. This limitation results from the non-random selection of schools, which depended on factors such as expressed institutional interest, proximity to the university campus, and the availability of technological infrastructure. Consequently, the final sample was heavily concentrated in two schools and included a relatively high proportion of students with mixed public–private educational backgrounds, limiting the generalisability of the findings.

Data collection between 2019 and 2023 was also significantly affected by the COVID-19 pandemic, which caused fluctuations in participation rates—most notably a sharp decline in 2020–2021—and led to the exclusion of schools without adequate infrastructure for hybrid learning. This, in turn, reinforced existing geographic and socio-economic imbalances in the sample. In operational terms, the absence of funding for student grants hindered the sustained participation of undergraduate women as peer models—a central element of the intervention—and prevented the completion of longitudinal assessments (one year or more post-intervention), limiting the evaluation of long-term persistence in STEM careers.

The analytical methodology imposed additional constraints, such as the exclusion of responses related to the military training track (CFO), which slightly restricted the scope of the career intentions analysis. Moreover, the study captures interest at a single point in time and cannot disentangle the project’s influence from deeper socio-cultural dynamics, including economic constraints and concerns about the ‘double shift’. Finally, the active involvement of one of the authors—although mitigated through peer review—introduces a potential proximity bias that cannot be entirely ruled out.

These limitations nonetheless provide valuable directions for future research. Longitudinal studies should be undertaken to track persistence and professional outcomes over time, supported by broader and more representative sampling that includes schools from less well-resourced regions. Further work should also explore socio-cultural and economic factors as potential levers for engagement in STEM. Securing dedicated funding to sustain peer-modelling initiatives and enable long-term evaluation would be crucial to transforming voluntary interventions into systematic, evidence-based programmes.