Analysis of the retention of women in higher education STEM programs

The paper addresses the persistent gender gap in STEM higher education despite global equity efforts. STEM is framed as applied and pure sciences (with medicine and health often included), with males typically reporting more interest in physical sciences and females in biological sciences. UNESCO data show low female enrollment in engineering, manufacturing, and construction relative to males, with variation across countries and exacerbation by stereotypes and cultural norms. The authors note a leaky pipeline across university entry, labor market, and leadership stages, and the scarcity of recognized female role models. The study aims to identify: (1) internal and external factors affecting women’s persistence in STEM at a private multi-campus Mexican university and their relation to Social Cognitive Career Theory (SCCT); and (2) actions to favor retention of women already enrolled in STEM programs.

A Scopus scan shows research on women’s STEM retention is relatively recent and grew post-COVID-19 but remains limited, especially on concrete reasons for leaving and intervention effectiveness. Theories of marginalization and validation suggest women are more prone to doubt their abilities and succumb to stereotypes; in the U.S., women are less persistent than men in completing STEM degrees (48% vs. 65%) despite similar preparation. Social factors (gender equity, life satisfaction), values, self-efficacy, and prior experiences influence decisions. Emphasizing STEM’s social orientation may attract more women. Interventions include multifaceted programs supporting psychosocial adjustment, milestones, and goal achievement, and sensitizing faculty to gendered dropout motivations and stereotypes. The literature distinguishes horizontal (recruitment vs. retention) and vertical (career progression) gender balance: female role models are especially important for recruitment; both male and female role models aid retention. Motivations involve parental influence, achievements, stereotypes, and attitudes (including spatial ability). Campaigns range from talks to interactive experiences and technology-mediated outreach, though retention interventions are less evaluated. Underrepresentation persists globally; factors include bias, harassment, lack of role models, stereotypes, lifestyle and work-family preferences, and field-specific beliefs. Despite improved job outcomes for female STEM graduates, representation gaps remain. The SCCT framework (self-efficacy, outcome expectations, goals) is presented as a lens for understanding interest formation, choices, and success/persistence in STEM.

Design: Three stages: (a) diagnosis using historical university data; (b) survey of undergraduate STEM students; (c) data analysis and proposal development.

Data source for diagnosis: University Data Hub (UDH) with anonymized records for enrollment, transfer, dropout, and completion across 26 campuses. Data available starting Fall 2014. Focus on School of Engineering and Sciences and School of Medicine and Health Sciences (data availability constraints). Outcomes: female enrollment share in STEM programs, female retention, and historical academic performance. Programs in Built Environment and Creative Studies with technological content were excluded due to data interface limitations.

Survey: Instrument grounded in SCCT to identify factors influencing retention and sense of community and cognitive career decision-making (interests, choice, success/performance). Content included demographics (entry year, gender, campus), career path (major changes), identification with the current major, prior STEM educational experiences (high school courses, extracurriculars), and open-ended prompts on attitudes of others, self-attitudes/values, external situations, and actions by faculty/classmates. Example mapping to SCCT factors provided (Table 1). Administration via Google Forms, March–June 2022, using snowball (random pop-up) sampling. Participation was anonymous with implied consent; no personal data collected. Ethics approval was obtained from Tecnologico de Monterrey.

Analysis: Descriptive quantitative analysis of survey responses; qualitative analysis of open-ended responses via stratified prospective analysis under postpositivist/critical theory approaches. Demographic and performance analyses from UDH. Theory of change used to draft retention proposals.

Institutional diagnostics (2014–2020):

• Total enrolled students: 77,517; women: 45% (34,703).
• STEM selection overall: 49% (37,984) of all students; among STEM students, women comprised 36% (13,675), meaning only 17% of the total university population were women in STEM.
• Slightly higher dropout in STEM among women vs. men: 8.60% vs. 8.35%.
• 75% of STEM dropouts had GPA ≥ 70/100 (passing), indicating academic performance did not primarily drive dropout.
• Highest dropout concentration in first four semesters, peaking in 2nd and 4th semesters.

Survey (n=49; post-COVID cohort; majority nearing end of 3rd year):

• All respondents took core high-school STEM subjects; about half of current STEM students reported extracurricular STEM activities (e.g., science workshops, robotics, lab experiences), which positively influenced STEM choice.
• Women highlighted inspiring, empathetic faculty as key to persistence; negative factors included competitive/toxic environments, overwhelming workload, and less empathetic or ineffective teaching.
• Passion/interest and identification with the career had the strongest perceived impact on choice and retention; external situations had medium impact. Family support and professional networking also contributed.
• Students in medicine cited altruism and research interest; non-STEM students cited creative/artistic interests. Few cited salary; one male mentioned it.

Coded internal/external factors from women’s responses (Table 6 counts):

• Internal: Self-perception (41), Interests (17), Personal experience (7), Self-efficacy (5).
• External: Faculty (27), Peers’ commitment (17), Extracurricular courses (6), Challenges/Stakeholders (4), Family support/communication (2), Curricular courses (2), Multidisciplinary (1), University environment (1), Networking (1).

Overall: Academic achievement did not explain women’s attrition; early semesters are most vulnerable; faculty inspiration/mentoring and prior extracurricular exposure are strong positive influences, while competitive climates and unsympathetic instruction are negative.

Findings align with SCCT: interest is nurtured by early exposure and personal experiences (extracurriculars, stakeholder interactions, information access, family support); persistence relates to self-efficacy built through successful academic experiences; and external environmental supports or barriers (faculty practices, peer climate, institutional environment) validate or undermine goals. The slight female–male dropout difference, concentration of attrition in early semesters, and the non-role of low GPA point to environmental and psychosocial factors rather than academic inability. Faculty behaviors (empathy, mentoring, inspiration) emerged as pivotal levers, while toxic competitiveness and depersonalization by instructors discouraged persistence. Early, socially oriented STEM experiences during high school can enhance identification with STEM and sense of belonging, which literature cites as a strong retention predictor. Based on these insights, retention strategies should target the first two years, enhance mentoring and sense of belonging, involve families, and equip faculty with training to mitigate biases, support diverse learners, and structure courses to increase predictability and perceived control.

The study confirms a significant gender gap in STEM enrollment and retention at the institution, with only 17% of the total student body being women in STEM and slightly higher female dropout than male despite adequate academic performance. Internal (self-perception, interests, self-efficacy) and external (faculty influence, peers, extracurricular exposure, family support, environment, networking) factors shape women’s choices and persistence, consistent with SCCT. The authors propose actions to enhance retention of enrolled women: remote and in-person mentoring programs, digital networking/support platforms with trained tutors and awareness workshops, and systematic faculty development. Monitoring indicators should accompany implementation to evaluate increases in women’s STEM enrollment and retention, especially focusing on the first two years. Future institutional work will extend mentoring models and report outcomes.

The study uses primarily descriptive statistics and limited qualitative coding; more rigorous quantitative analyses and deeper coding are needed. It was conducted at a private university, potentially limiting generalizability; access constraints and socioeconomic factors may differ from public institutions. The survey sample was small (n=49) and collected during a post-COVID transition, possibly affecting participation and experiences. Differences in dropout rates by high school background (public vs. private) suggest contextual effects. Some challenges (e.g., adapting to faculty and peers) were not exclusive to women; future studies should examine gender-diverse experiences, including non-binary students, in greater depth.