Determinants of Women’s Persistence in Engineering: A Structural Equation Modeling Study in Universities in Northern Peru (#2190)
Read ArticleDate of Conference
July 15-17, 2026
Published In
"Engineering without Borders: Artificial Intelligence, Knowledge, Innovation, and Alliances for a Future from the Americas"
Location of Conference
Santiago (Chile)
Authors
Minchola Vasquez, Angelica Maria
Medina Marchena, Agustín
Olaya Guerrero, Lidia Mercedes
Minguillo Rubio, César Augusto
Corrales Otazú, Christian David
Apaza Miranda, Sarita Jessica
Li Vega, Fiorella Vanessa
Abstract
This study aimed to determine the effect of perceived teacher support and classroom climate on persistence among women in engineering programs, assessing the mediating role of academic self-efficacy through a structural equation model in universities in northern Peru. A quantitative, explanatory, non-experimental, cross-sectional design was employed with a sample of 350 female engineering students. Four reflective latent constructs (teacher support, classroom climate, academic self-efficacy, and persistence) were measured using 16 Likert-type items. The measurement model showed high reliability and adequate convergent and discriminant validity, while the overall model fit was excellent (CFI=0.975; TLI=0.968; RMSEA=0.048; SRMR=0.035). Structural results indicated that teacher support and classroom climate positively and significantly predicted academic self-efficacy (β=0.42 and β=0.31), and self-efficacy, in turn, predicted persistence with a substantial effect (β=0.48). Additional direct effects from teacher support and classroom climate to persistence were also significant, and indirect effects confirmed partial mediation via self-efficacy for both contextual predictors. The findings suggested that pedagogical context fostered persistence through motivational pathways and complementary direct influences. Recommendations included strengthening faculty development in actionable feedback and academic encouragement, promoting equitable and psychologically safe classroom climates, and extending evidence through longitudinal designs and subgroup analyses.