Predictive Model for STEM Vocational Guidance through Profile Analysis and Information Adaptation with a Gender Perspective

Abstract

This article analyzes a predictive model for vocational guidance focused on identifying student profiles toward STEM careers (Science, Technology, Engineering, and Mathematics) with an explicit gender perspective. The study is based on data from students at University X, considering academic variables such as standardized test results, socioeconomic background, and gender markers. Data preprocessing techniques, cluster analysis, and dimensionality reduction through Principal Component Analysis (PCA) were applied to identify significant patterns in academic behavior. The results allow the classification of students into specific profiles, highlighting those with skills aligned with STEM careers, cases where the model presents accuracy limitations, and differentiated trajectories mediated by gender. Importantly, the analysis reveals that, although female students demonstrate strong competencies in areas such as Mathematics, Critical Reading, and Natural Sciences, external factors such as socioeconomic conditions and gender roles influence their vocational decisions. By integrating this gender-sensitive analysis into the predictive framework, the study establishes a solid foundation for adapting information in future model iterations, improving the accuracy of inferences, and promoting vocational guidance processes that are not only personalized and adaptive but also explicitly oriented toward equity in STEM participation.