Why STEM feels like a ‘non-place’ for women? What schools can do about it

Despite decades of gender advocacy, STEM fields continue to sideline women, with information and communication technologies (ICT), engineering, and artificial intelligence (AI) remaining some of the most male-dominated domains. A newly published study in Social Sciences titled “STEM, a Non-Place for Women? Evidences and Transformative Initiatives” highlights not only the persistence of the problem, but also unveils a pioneering educational model that seeks to dismantle structural barriers and reimagine how STEM is taught to girls and young women.

Why do STEM fields remain unwelcoming for women?

According to the study, systemic gender bias still saturates STEM education and professional domains, reinforced by a male-dominated culture, institutional inertia, and AI systems that perpetuate inequality. Women account for just 32.8% of STEM graduates across the EU and are especially underrepresented in ICT, where they make up only 17.2% of students in Spain. The study draws on Marc Augé’s theory of “non-places” to describe STEM environments as devoid of female belonging, relational context, or identity formation, spaces where women remain outsiders despite formal inclusion.

Numerous institutional reports, including those by UNESCO and Eurostat, confirm these trends and add that women make up a fraction of ICT patent applicants, AI start-up founders, and academic staff in AI-related fields. The authors link this alienation to broader issues such as the sexualization of women in generative AI systems, exclusion from videogame culture, and a lack of critical pedagogy in AI education.

The report also critiques large language models, such as GPT and LLaMA, for encoding and reproducing gender biases. These systems, frequently used in educational tools, have been found to exhibit stereotypical portrayals of gender roles and exclude women from algorithmic visibility. The study stresses that unless educational content directly confronts these biases, AI will reinforce rather than disrupt systemic gender hierarchies.

What are Spanish universities doing to promote gender equity in STEM?

Despite a robust legal framework that includes Spain’s Organic Law 3/2007 on gender equality and the presence of 54 universities in the RUIGEU network for gender excellence, institutional resistance remains high. While most universities have equality plans and dedicated gender units, implementation gaps persist, particularly in STEM education.

The study maps out various national initiatives: mentoring schemes like Inspira STEAM and FEMenGin, gender mainstreaming guides for university teaching, and targeted interventions such as Catalonia’s integration of gender-related competencies in quality assurance for higher education. Despite these initiatives, female visibility and gender-sensitive teaching remain weakly embedded in computer science, engineering, and AI courses.

One standout example is the University of Santiago de Compostela (USC), which has been at the forefront of transformative initiatives. Since 2010, it has awarded grants for embedding gender perspectives into research and teaching, and its program “Unha enxeñeira ou científica en cada cole” (“One female engineer or scientist in each school”) has reached nearly 200 schools across Galicia. The program emphasizes female role models and provides hands-on STEM workshops for primary students, laying the groundwork for more inclusive early exposure to STEM careers.

How is AI and computational thinking being reimagined for equity?

The study’s most compelling contribution is the design and pilot deployment of a groundbreaking educational toolkit titled A informática e a vida (“Informatics and Life”). Developed by CiTIUS at USC, the program comprises ten computational thinking (CT) and AI-related activities tailored for students aged 9 to 15. It focuses not on coding literacy per se, but on conceptual understanding, ethical reflection, and gender-inclusive problem solving.

The activities deliberately avoid programming tools like Scratch, which the researchers argue reinforce gendered stereotypes due to their videogame-like aesthetics. Instead, they introduce Python in a storytelling framework and draw on analogies with biology, medicine, and social sciences to make CT more relatable for girls.

For instance, activities like “Who invented computing?” recover historical contributions by women in computer science, while others model AI decision-making through real-life scenarios such as spaghetti sorting and ball kicking to explain machine learning logic. One activity helps students identify gender bias in AI by demonstrating how skewed training data can lead to exclusionary outcomes. All exercises avoid data collection, respecting students’ privacy and reinforcing human-centric AI design principles.

The initial pilot in February 2024 attracted just 15 schools, but interest surged following public symposiums and teacher training workshops. By March 2025, over 30 educational centers in Galicia had enrolled, indicating strong demand for gender-sensitive digital literacy programs. Teachers responded positively, with satisfaction scores averaging above 4 out of 5, praising the program’s innovative, non-commercial approach and its effectiveness in building conceptual clarity and inclusive learning environments.

Toward a more inclusive future in STEM education

The study urges expansion of gender perspective across all levels of education, from early childhood through university. It argues that empowering girls to see themselves as creators, not just users of technology, requires a radical shift in curricula, teacher training, and pedagogical tools. The proposed model exemplifies how CT can be introduced without defaulting to male-coded tech tropes, thereby creating a sense of belonging for girls in spaces historically structured to exclude them.

Plans are underway to scale Informatics and Life by collecting formal feedback through new teacher questionnaires and refining activities based on classroom implementation. Researchers also propose incorporating male allies as role models to promote gender-equitable behavior among boys.

The study makes it clear that no amount of institutional policy will meaningfully close the gender gap in STEM unless it is matched by pedagogical reform that reshapes how knowledge is taught, who is centered in its narrative, and what kind of technological future we choose to build. Only by dismantling the invisible codes of exclusion embedded in the tools and cultures of STEM can we hope to transform these “non-places” into truly shared spaces of innovation, equity, and inclusion.